Skip to main content

Full text of "He shu Dong Yindu Hua qiao shang ren"

See other formats

Go ogle 

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 
are our gateways to the past, representing a wealth or history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are aoing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright inrringement liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 
at http : //books . qooqle . com/] 

Digitized by 






飞 、、モ や • 

Digitized by 








From the Proceedings of the United States National Museum, Vol. XVI, pages zxQ-at6 
(with Plates XIX-XXXV). 

[No. 930.] 



Digitized by 

Digitized by 


' , 一 

に 一、 • " ― , ' ^ 

ムし V:? 

Scientific results of explorations by the U. S. Fish Commission Steamer Albatross. 

[Published by permission of Hon. Marshall McDonald, Commissioner of Fisheries.] 

OF 1887-1888. 

* RY 

J. Playfair McMurrich, M. A., Ph. D. 
[With Plates xix-xxxv.] 

The collection which forms the subject of this report was forwarded 
to me soon after its arival in Washington, and I gladly availed myself 
of the ODportunity thus presented of continuing the investigation of 
the deep-sea Actiniaas, which was so admirably inaugurated by Prof. 
Eichard Hertwig. The studies of this distinguished naturalist have 
resulted m the establishment of a new and 】nore correct basis for the 
classification of the Actiniae, by calling to the aid of the somewhat un- 
certain external peculiarities, the more reliable characteristics revealed 
by a tUorough anatomical study of each species. The revision of the 
Actinians in accordance with this new system of classification founded 
by Prof. Hertwig has been carried on by myself for the Actiniae of the 
West Indies and by Prof. Hf^ddon for the forms occurring on the coasts 
of Great Britain. Much has been added to our knowledge of many 
forms, and many errors have been corrected, and it has been my hope 
that the present study would clear away still further the mists that ob- 
scure the relationships of the various Actinian groups. 

The present report deals with tlie Edwardsise, Protactinia?, Hexac- 
tinife, and Ceriantheae obtained by the Albatross, I hope in a future 
report to give the results of my studies of the Zoaiithcfe. 

I gladly acknowledge the many courtesies I have received from my 
friend, Mr. Eichard Eathbuu^ during the preparation of this report. 
I am indebted to Mm for the opportunity of comparing several specimens 
in the collection with allied and occasionally identical forms obtained 
by the Fish Commission steamers off the eastern coast of North America. 

Part I. 


What may be termed an approximately correct idea of the relation- 
ships of the various groups of animals now included under the term 
Authozoa or Actinozoa can be said to have come in on】y with the begiii- 

Proceedings National Museum, Vol. xvi— No. 930. 


Digitized by 


niDg of the present century, and to have had its first exponent in Cuvier. 
Earlier authors were led astray by the supposed vegetable character 
of the corals and similar formSy and later, by attaching too great im 
portance to the presence or absence of a hard skeleton, whereby closely 
related forms were widely separated. Thus Liniie in the twelfth edi- 
tion of bis " Syatema" referred the genus Actinia to the Mollusca, the 
remaining Actinozoa being referred to two groups, the Lithophytes, 
which included the Madrepores, and Zoophytes, which, in addition to 
the Alcyonaria, contained also sponges, Bryozoa, Sertularia, and Pro- 
tozoa ( Vorticella). Pallas (1766) improved this arrangement slightly 
by fusing the Lithophytes and Zoophytes to a single group, but the 
genus Actinia he referred, along with the Ecbinoderms, to bis group 

Cuvier by the foundation of the Radiata, a group containing, it is 
true, very heterogeneous members, did good service in bringing to- 
gether more closely than previous authors the allied Anthozoa. The 
third class of the Kadiata, the A.caleplis, contained the genera Actinia 
and Zoanthus with which was associated Lucemaria, while in the fourth 
class, that of the Polypes, were grouped together the rest of the An- 
thozoa under the term "Polypes corticaux a poly piers." The tribes of 
this latter group with some of their principal genera are as follows : 

Tribe 1 CersLtophytes—AntipatheSj Gorgonia. 

2 Lithophytes 二: ム, •*, Madrepora^ Millepora. 

3 Polypes usigenrs^Pennatula^ Rmilla, to which were added 

4 Alcyou8=Alcyoniumj Spongia. 

It will be seen that the character of hardness or softness was given 
considerable weight in the Cuvierian system, leading to the association 
in the same tribe of an Alcyonariaii, a Hexacoiallian, and a Hydrozooii, 
and similarly to the separation of various Alcyouarian genera, accord- 
ing to their relative consistency. The separation of the Actinians 
from the Mollusca and their reference to the Acalephs is however a step 
in advance, though their true relationships were unpeiceived. 

With contemporary and succeeding systeinatists these two features 
held firm ground. Lamarck (1818) though referring Zoanthtis to the 
Polyps with Hydraj Coryne, etc., returns to the classification of AcUma 
with the Ecbinoderms as advocated by Pallas, being followed in this re- 
spect by Schweigger ( 1820), who makes the presence or absence of a hard 
skeleton the criterion according to which the Zoophytes are referred to 
the Z, monohyla or Z, heterohyla, the former division eoutaining Infusoria, 
Eotifera, ZoanthuH, Tulmlaria, and the Alcyonicls. The reference of the 
Actiuians to one of the groups of the Polypes dates back to Lamou- 
ronx (1821), who still relying on the presence or absence of a skeleton 
divides the Zociphytes into (1) Polypiers flexibles, (2) Polypiers pier- 
reux, and (3) Polypiers sarcoides, the last group containing the Ac- 
tiuians together with the Alcyonids and the compound Ascidians. 



Notwithstanding the heterogeneity of these groups, Lamouroux's 
classification paves the way for the more accurate systems, that fol- 
low. ! N^oticeable especially is that of de Blainville (1834), who asso- 
ciates together in Class III Zoanthaires ot his Type Actinozoaires the 
Actinians, Zoanthans and Madrepores, thus cutting loose from the 
consistency systems of his predecessors. The remaining Anthozoa, 
together with the Hydroids, Millepores, and Bryozoa, he refers to the 
fourth class, Polypiaires. 

Before de Blainville, however, Eapp ('29) had published a classifica- 
tion of the Polyps which, though not accepted by his successors, stands 
out, in the light of our present knowledge, as an evidence of the value 
of anatomical distinctions as a basis for classification. In his preface 
Rapp says: "Bei dem Stiidien der mit cinem Geriiste oder Poly pen- 
stock verseheneii Polypen war dieser Theil, iiideui man das Thier selbst 
vern aclilassigte, bisher hauptsaclilicher Gegeustand der Aufmerksam- 
keit. * * * Zwar felilt es iiber diese Thiore iiieht an trefflichen 
Beobachtungen, welche man haiiptsaclili(»b der iieiiesten Zeit verdankt, 
aber sie stelieii bis jetzt meist iioeh zii isblirt, als dass sie auf die ganze 
Gestalt desjeiii^jen Theils der Wisseiiscbaft, welcber mit diesen Ge- 
schopfen sich beschaftigt, einen durcligreifendeu Einfluss gehabt hal- 
teii." To bring these isolated anatomical facts together, and to add 
to them was the task Rapp set himself, and as the result of liis- studies 
two important facts were brought to light. In the first place he recog- 
nized the near relationship of the Madrepores and the Actinians, and 
secondly he discovered the Actiniaii nature of the form previously 
described by biin as Tubular ia solitarla, now kuowii by the generic 
name of Cerianthus proposed by Delia Chiajeiu 1832. 

RaBp assumed as the basis of his classification the mode of forma- 
tion of the reproductive organs. He found that some polyps produced 
ova on the outer surface of the body, while in others tlie "Keimkorner" 
had their origin in the interior; the former constitute his Exoarier, 
while the latter are referred to tlie division Endoarier, To the former 
division he assigned the Hydras, Corynes (including Sertularia and 
Tuhularia) and Millepores (! ), while to the latter were referred the 
Alcyonids, Tubipores, Corals (a group which included (Jomlli 請, Gor- 
gonia, Isis, and AntipafhesC^)*), Pennatulids, Zoanthids, Madrepores, 
and Actinians. Bearing* in mind the fact already stated that Eapp 
associated the forms now known as Cerianthus witli the Actinians, it 
may be seen that liis division Endoarier is equivalent to the modern 
group Antliozoa, while his Exoarier corresponds essentially with the 
Hydrozoa, though he does not include within it tlie Hydroiiiedusse, 
whose relationships to the Hyclroids had not been discovered. 

It is interesting to note that in this classification Eapp forestalled the 
Hertwigs (,79), whose proposed division of the Coelenterates into 

The mark oi interrogation is Rapp, 8. 


Ectocarpae and Endocarpae is founded on the identical characteristic 
which Eapp chose, tliough the more receut authors define more accu- 
rately the place of origin of the reproductive elements in the terms of 
the germ-layers, structures unknown to liapp. 

Unfortunately, the systematists who immediately succeeded Rapp did 
not advance the position be liad occupied. ])e Blainville's association 
of the Actinians with the Madrepores has already been noticed, a liai)py 
exception to the complicated confusion into which he falls as to other 
groups. On the whole, however, his classification must be considered 
an advance as compared with that of Ehreuberg ('34), who falls back 
to the old consistence system, though avoiding l)e Blaiiiville's perpet- 
uation of the earlier misconception of the Bryozoa as allied to the Zoo- 
pliytes. To Ehrenberg* we owe the substitution of tlie term Autliozoa 
for that of Zoophyta, employed by earlier writers, and this "circulus" 
he divides into two orders whose limitations may be seen from the fol- 
lowing synopsis: 

Circulus I. ― Antbozoa. 
Ordo I. 一 Zoocorallia. 

Tribus I. ― Zoocorallia polyactiuia (Actiuiaus, Zoanthans, and Fuugiclte). 

Tribus II. ― Zoocorallia octactiuia (Xeuian^ Tubiporids, Alcyoiiids, and 

Tribus III. ― Zoocorallia oligactinia (Hydroids). 
Ordo II. ― Phytocoralli.i. 

Tribus IV. ― Pbytocorallia polyactiuia (Oculinids and Astraeids). 

Tribus V. 一 Pbytocorallia dodecactinia (Madrepores and Millepores). 

Tribus VI. ― Phytocorallia octactiuia (Corallium, Isids, and Gorgon ids). 

Tribus VII. ― Phytocorallia oligactinia (Jllojwra). . 

It will be seen from the above tliat the Zoocorallia includes all those 
forms wliicli arc destitute of a hard skeleton, or which, like Fungia, 
possessing a corallum arc not fixed, while the Phytocorallia embraces 
the forms provided with a hard skeleton, being at the same time 
fixed. Such a elassificatioii necessarily separates closely allied forms, 
as, for instance, the FuDgida^ from the other Hexacorallia, and the 
Pennatulids from the Gorgoiiids. The group Anthozoa as conceived 
by Ehrenberg differs from the modern conception of the group iu in- 
cluding the Hydroids and Hydrocorallinje, in which respect Ehrenberg 
falls far behind Eapp, and in excluding the Aiitipatharia wliicb are iu 
this system referred to the Bryozoa. In one particular, however, 
Ehi euberg surpasses liis predecessors, witli tlie exception of liapp^aud 
that is in employing for his secondary groups characters which belong- 
to the living animals. The number of the tentacles is a feature which 
within certain limits lias been found to be associated with the features 
which mark out the various groups as now recognized. 

The association of the various eight-teiitacled forms into a single 
group was one of the important steps whicli now followed. According 
to a statement made by Dana ('4Ga) this was first done by Milne Ed- 

Digitized by 


wards,* who divided bis group of the Polypes parencbymates into three 
groups : 

Sertulariens. Zoanthaires. 

Of these the first group corresponds to the Hydroidea, the second to 
the Actiuiaria and Hexacorallia, and the last to the Alcyoiiaria. 

A most important classification appeared in 184(3 as the result of the 
extended study of the Zoophytes of the Wilkes exploring expedition 
by Prof. James D. Daua. His groups are as follows : 

zoOphyta. • 

I. Order. Actiuoidea. , 

I. Suborder. Actinaria. 

I. Tribe. Astrieacea— including the Actiniaiis with which Lucernaria was 
' associated and the Astreid and Fungi" corals. 

II. Tribe. Caryophyllacea— including besides the Caryophyllids and Cyatho- 

phyllids, the Zoauthea*. 

III. Tribe. Madreporacea 一 including Madreporida, Favositids, to which are 
referred the Milleporcs and Pod t ids. 

IV". Tribe. Antipatbacea. 

II. Suborder. Alcyonaria. 

II. Order. Hydroitlea. 

It will be seen from this that the order Actinoidea is practically 
equivalent to the group Aiitliozoa of to-day, and that a clear distinc- 
tion is made between tlie Actiniaria and the Alcyonaria. The former 
group includes all the Hexacorallia and the Actiniaria of later authors, 
as well as the Antipatliaeea, ami it is interesting to note that Dana in- 
sists upon the unimportance of tlie stony coralliim, grouping together, 
as De Blainville had done before him, the non-skeletogenous Actinians 
and the skeletogenous Hexacorallia. 

One of tlie principal groups of the Aiitliozoa, the Alcyonaria, being 
thus delimited, and a second, the Autipatliaria, also marked out, though 
not considered of equal value, it will be well to go back some distance and 
note the gradual discovery of various forms recognized now as distinct 
groups, but included so far as known in the first two tribes of Dana's 

The earlier authors recognized ji single genus of Actinia only, though 
other names, ~ e, Urtica, Hydra, and Priapus —— lind been proposed. 
Ill 1801 Lamarck separated the genus Zoanihm for the form described 
by Ellis as Actinia ftociata, and thus paved the way for the distinction 
which later authors made between this and similar forms and the 
Actiniae proper. Cuvier also recognized the genus Minyas, referring 
it, however, to the Holothurians, its true position not being recognized 
until later by Lesueur ('17). A further division of the genus Actinia 
was inaugurated by Oken in 1815, who established the genera Metrid- 
ium and Cereus, and set the example for the more accurate generic 

*I take this statement from Dana, not having access to Milne Edwards' work. 

Digitized by 


classification found in later authors. The large number of forms 
brought to notice by the scientific voyages of this period increased 
noticeably the number of Actinian genera, and in the classifications of 
De Blamville and Ehrenberg we find a considerable number of genera 

Attention has already been called to the discovery of the Actinian 
character of Gerianthus by liapp ('29), the subsequent application of 
the generic name by which it is now known by Delia Chiaje. In 1841 
Quatrefages ('41), in a paper which is a model of accurate observation 
and description, established the genus Udwardsia on essentially the 
same basis as that on wbicli it now rests, though more recent obser- 
vations have added certain particulars which the methods of micro 
scopic investigation of the day liave brought to light. 

The year 1841 marks therefore the establishment of most of the groups 
of Aetinaria which are now recognized, so far as they possessed generic 
value, but for some time Cerianthus atid Edwardsia were considered of 
equal taxonomic value with Actinia^ Thalassianthus^ JDiscosoma^ and 
other simply generic terms. The Zoanthm group formed to some extent, 
however, an exception to this rule, probably on account of their colonial 
habit of life and the power some possessed {Palythoay Corticifera) of 
encrusting themselves with calcareous or siliceous particles, recalling 
by their consistency skeletogeaous forms. Their geminiparous repro- 
duction induced Dana to group tliein apart from the rest of the Actin- 
iaus, and associate them with the Oaryopliyllid corals. 

To give a resume then of the state of Antliozoan taxonomy at the 
middle of this century it may be said that the group was definitely 
delimited, the Bryozoa having been excluded in accordance with the 
observations of Milne-Edwards. The Alcyonarian forms bad been 
grouped together from their earlier separation into a number of groups 
each equivalent to the Hexacorallia, Actiniae, etc. The Antipatliaria 
were referred to tlie Anthozoa, and even constituted a group of slightly 
less value than the Alcyonaria. And lastly, tlie Actiniae had been 
divided into a number of genera and associated with the Hexacorallia, 
tbe similarity of structure of the animals themselves being considered 
of greater moment than the possession or absence of a corallum. 

A new era in Antliozoaii classification was introduced by the publi 
cation in 1857 of the first two volumes of Milne-Edwards' Histoire 
naturelle des Coralliaires, In some respects, notably in tbe severance 
of the Madrepores from tbe Actiuiana, a backward step was taken 
which has been maintained up to a comparatively recent date, but on 
the other hand, a decided advance was accomplished in the more accu- 
rate delimitation of several groups, and in the recognition of groups of 
genera among the Actiiiiaiis. 

Milne-Edwards recognized Leuckart's division of the Cuvierian Radi- 
ata into the two groups Ecbiiiodermata and Coelenterata, and divided 
the latter into two classes, the Acalepbs, including three sub-classes, 



L e" the Medusae, Sipbonophores, and Hydroids, and the Ooralliaires. 
The Ooralliaires he again subdivided as follows : 
Class Ooralliaires. 

Sub-class Cnidaria. 
Order Alcyoiiaria. 
" Zoantliaria. 
Sub-order Zoantharia malacodermata or Actinaria. 

Zoaiitbariii sclerobasica or Antipathaiia. 
ZoaDtbaria sclerodermata or Madreporaria 
Sub-class Podactinaria (=Lucernaria). 
It will be seen from this that Milne-Edwards's class Ooralliaires is 
equivalent to Dana's order Act ino idea, and his sub-class Cuidiiria to 
liana's Actinaria mirnis Lucernaria, a step toward the separation of this 
genus from the Anthozoa, and its reference to the modem group of the 
Scyphozoa. In his division of the Zoantharia, however, Mihie-Edwards 
retrogrades towards the older consistence systems of Lainouroux aud 

So far as the Actinaria are concerned Milne-Edwards did excellent 
service in delimiting' tke various genera that bad been proposed, in 
dividing these up iu some cases, and estabiisliiiig new genera, such as 
FaractiSy PhyniactiSj Oulactis, etc., and in grouping* similar genera to- 
gether, forming families, sub-families, etc. His larger divisions are as 
follows : 

1. Family Actinidte. 

1. Sub -family Miuadae. 

2. Sub-family Actiuinff . 

3. Sub-family Tbaliissiaiitbinje. 

4. Sub-family Pbylhu'tiua*. 

5. Sub-family ZoanthinSe. 

2. Family Cerianthidae. 

The sub-family Actiiiin^e was again subdivided into sections, thus: 

1. Actinines vulgairos ― including forms with smooth walls and adherent base. 

2. Actinines verruqiiciises 一 including fornus with tuhorclos or verrucaB upon the 


3. Actinines perforees 一 corresponding to the family ト;; igartidae of id ore recent sys- 


4. Actinines pivotantes ― including forms which do not possess an adherent base. 

In analyzing this classification in the light of our present know レ 
edge of the relationships of the Authozoau groups we note n recogni- 
tion of most of the niodern taxouomic groups, with, however, very 
rmeqiia】 values attached to them. Thus the Alcyonaria constitute 
a group equivalent to all the others taken together; the Autipatharia, 
another of equal value with all that still remains; the OeriantliidyB 
are recognized as a family equal in value to all the other Actiniaiis; 
while the ZoantbinaB are equivalent only to the Tbalassianthiiue, etc. 
Tbe Edwardsiae do not liave a group value, being recognized simply as 
a genus of Actinines pivotantes, where they are associated with Ilyan- 
thusj Pemhia, and Sphenopiis, the last belonging properly to the Zoan- 

Digitized by 


thinae. In choosing the relative consistency of the various forms as a 
basis for his division of the Zoantharia Milne-Edwards naturally falls 
into certain of the errors which such a classification entailg, and which 
had been handed down from earlier days, as for instance the group- 
ing of the Millepores with the Maflreporaria. In this, however, there 
is neither loss nor gain, since none of his predecessors, with the con- 
spicuous exception of Eapp, had suggested the reference of these forms 
to their proper position. The principal error of the classification, as 
already pointed out, lay in the attaching of too great importance to 
the presence or absence of a corallum, and iu the disregard of the sim- 
ilarity of the soft parts of Madreporaria and Actinaria so definitely 
stated by Dana. 

Milne-Edwards's classification had a marked influence upon later 
writers, most of whom adopted his larger divisions, the principal mod- 
ifications introduced by them affecting the arrangement and definition 
of the lesser groups. An exception to this, however, was the classifi- 
cation of Gosse (,60) who adhered to the arrangement laid down by- 
Dana, but went a little farther than that author in dividing certain of 
the tribes of Actinaria into families, thus: 

Suborder-Actinaria ― 
Tribe I. Astraeacea : 

Family I. Metridiadae = forms with compound tentacles. 

II. Sagartiadae = with simple tentacles, adherent base, and column 
pierced by cinclides. 

III. Antheadae = col amn smooth and imperforate, margin simple. 

IV. Actiniad8B= margin beaded. 
V. Bunodidae = column warted. 

VI, llyanthidfB = base non-adherent, rounded, simple. 
VII. Minyadid{B=ba8e non-adherent inclosing an air chamber. 
Tribe II. Caryophylliacea : ' 
a. Without a corallum. 
Family I. C apnead ― simple. 

II. Zoanthida> = compound. 
p. "With a coralliim, certain corals divided into four families. 

The coralligenoiis Astraeacea Gosse does not classify, none of the gen- 
era being British, nor does he divide the Madreporacea or Antipathacea 
into families, for the same reason. The Lucernariadae he excludes from 
the Actinaria, recognizing their affinities to tbe Medusae. ' 

On comparing this classification with that of Milne-Edwards, it will 
be seen that, independently of the association of noii-coralligenous and 
coralligenous forms, there is a very different grouping of the genera. 
Tbe family Metridiadie (a name badly chosen) is equivalent to Milne- 
Edwards's ThalassianthiD<ne and Phyllaotinae, the Sagartiadje are the 
Actiniiies perforees, and the BnnodidaB the Actinines verruqueuses, 
both raised to the rank of families. The Actinines vulgaires are 
divided into three families, two of which, the Antheadse and Actiniadae 
belong to the Astraeacea, while the third, the Capneadae, is referred, 
with the Zoanthidae, to the Caryophylliaceae, while Milne-Edwards's 

Digitized by 


family Oerianthidae is abolished, Cerianthus and Arachnactis being 
associated with his Actiniues pivotantes to form the family IlyanthidaB. 
These comparisons refer to the broad features of the groups, there being 
differences in detail in some. Many new genera were established by 
Gosse, as for instance, Bolocera, Bunodes and Aiptasia, and this, as 
well as his disregard for the most part of non-British forms, renders it 
difficult to make a detailed comparison between the two authors. 

By the exclusion of the LucernariadaB the Anthozoa obtained the 
limitations which they now possess, except that the Hydrocoralliaes 
still continued to be referred to the group. Agassiz indeed upheld their 
hydroid character, but it was not until Moseley's brilliant observations 
('78) were made, that they were definitely assigned to the position long 
before pointed out for them by Eapp. 

As already stated, subsequent authors were more influenced by Milne- 
Edwards than by Gosse in drawing up their classifications, though 
the division into smaller groups was not unlike that proposed by the 
latter. Gosse's smaller divisions were more or less adopted and sub- 
ordinated to Milne-Edwards's system. It will be altogether unneces- 
sary to refer to all the classifications presenting these features, but still 
it will be convenient to give one or two examples, choosing those which 
present most historical value. 

Oue of these may be the classification proposed by Verrill in 1865, 
which outdoes even that of Milne-Edwards in placing inordinate im- 
portance upon the corallum. Verrill divides the Cnidaria or Polypi 
into 3 orders, i. e., (1) Madreporaria, (2) Actiiiaria, (3) Alcyonaria, the in- 
crease from the number proposed by Milne-Edwards being accomplished 
by raising the Madreporaria from the subordinate position they occu- 
pied in the order Zoantharia and making them of equivalent rank with 
the Alcyonaria. The (livision of the Actiuaria which Verrill proposed 
was as follows: 、 

Suborder I. Zoanthacea. 

Families. ZoanthidtB and Bergidie. 
Suborder II. Antipathacea. 

Families. Antipathidte and Gerardidip. 
Suborder III. Actinacea. 

Families. ActinidsBjThalassianthidse, Minyidip, IlyanthidjB, Cerianthidse. 

This arrangement is important in giving the Zoanthids a greater 
importance than liad hitherto been assigned to them, and in separating 
the Oerianthidae from the Ilyanthidae, though they do not receive the 
same position that Milne-Edwards gave tbem. The family Actinidae 
Verrill divided in various subfamilies, differing somewhat from the 
equivalent groups of Gosse and Milne-Edwards, a subfamily Phellinae 
being established for tlie genus Phellia. Milne-Edwards' Pbyllactinre 
and Thalassiantliiiise he unites together in his family Thalassianthidfe, 
which is subdivided into the subfamilies Phyllactinae, Thalassianthinje, 
Heterodactylin^e and Discostominae (Verrill, '68), the members of 

Digitized by 


the 】ast named subfamily having the tentacles arraiiged in radiating 
rows, more than one tentacle communicating with an intermesenterial 
space. The establishment of this peculiarity is important, as it is a 
character which approximates the Actiiiaria with the Madreporaria. 

To the Discostominae Verrill referred the genera Biscosoma and Cory- 
nactis, classed by Milne-Edwards with the Actiniues vulgaires, and 
Capnea and Aureliana, referred by Gosse to the Caryopbylliacea. 

The classification of Klunziiiger ('77) may now claim our attention 
as showing a further step toward a correct differeutiation of the groups. 
The classification is to a very large extent similar to that of Verrill, but 
contains certain important innovations. The Madrepores are, following 
Milne Edwards, considered a separate group, and tbe remaining groups 
are as follows : 

I Or. Alcyouaria. 
II Or. Aiitipatharia. 

III Or. Zoaiitbaria ― including the Zoanthidse. 

IV Or. Actinaria. • 

1. Family Actinidsr. 

1. Subfamily Actininte. 

2. Subfamily Phellina>. 

3. Subfamily 8agartinie. 

4. Subfamily Bunodiute. 

2. Family Ilyauthifhf^. 

3. Family Cerianthida*. 

4. Family Discosomidse. 

5. Family Thala88ianthi(l«\ 

1. Subfamily PLyllnctina?. 

2. Subfamily Thalassianthiii*. 

The first noticeable feature of this classification is the separattou of 
the Autipatbaria and Zoautharia from the Actinaria, and the elevation 
of their rank to that of groups equivalent to tbe Alcyouaria. Further- 
more, among the lesser groups there is the separation of the Discoso- 
midae from the Thalassianthidjie, with which Verrill associated them, 
the radiate arrangement of the tentacle being the characteristic feature 
of the family. Kkmzinger, however, failed to associate tlie genus Covy- 
nactis with Dlftcomma, stating: definitely tbat its tentacles alternate 
with each other. The other families and subfamilies are essentially the 
same as those of Verrill, except that uo mention is made of tlie Miny- 

We come now to tlie moiiograpb of the Actiniaria by Andres ('83), 
wnich must ever remain a monument to the industry of its author, to 
wnom all actinologists are indebted for placing in their hands sucli a 
carefully collated aud complete list of the Actinians known up to 1880. 
Unfortunately for our purpose, Andres does not express his ideas as to 
the relationships wliich his Actinaria bear to the Aleyonaria and An- 
tipatharia, but confines his attention solely to the Zoantharia mala- 

)L. XVI, "1 

1893. J 



codermata of Milije Edwards. He divides the group into seven fami- 
lies, thus: 

Thalassianthin^e . 

The names of the majority of these groups indicate their limitations; 
the greatest innovations are the separation of the Edwardsias from the 
Actininje and the establishment of the Stichodactylinie. This family 
possesses for its distinguishing character the feature upon which Klun- 
zinger based his family Discosomidae, i. e" the radiate arrangement 
of the tentacles, but at the same time it is made much more comprehen- 
sive, the Phyllactinse of Klunzinger being associated witk Discosoma, 
Capnea, Aureliana. Phymanthus and other genera, all of which possess 
radially arranged tentacles. The Thalassianthinae is, consequently, 
poor in genera compared with Klunzinger's Thalassianthidae, contain- 
ing ouly a few forms with large compound tentacles. Four of Andres' 
groups are certainly well established, namely, the Edwardsinte, Ac- 
tininsB, Zoanthiuae and CerianthiDte. He was influenced, however, too 
much by the arrangement aud structure of the tentacles in making the 
Sticbodactyliuae and Thalassiantbinfe equivalent to these four; they 
should more properly be made subgroups of the Actininae. The same 
remark applies, perhaps, to the Minyadinse, though we are still in ignor- 
ance as to the structural peculiarities of its members. The fact that some 
of the species evidently have their parts arranged on a hexamerous plan 
favors this view, aud the occurrence of others possessing a decanierous 
arrangement cau not be considered as of great weight in favor of keep- 
ing them distinct, in view of the same symmetry occurring in the Hal- 
campidse, for instance, and in other sporadic instances in which there 
can be no question as to the advisability of associating the forms with 
their hexamerous relatives in the Actininae. In fact, it seems proba- 
ble that the MinyadinaB are not even to be given a value equal to the 
Stichodactylinie, but are rather to be referred to the Halcampidae, a 
family or subfamily of the Actininae. 

The disappearance of the Ilyanthidae from the list of families is an 
important point also. Andres lias diminished the importance of trivial 
characters in accomplishing this, and lias emphasized the importance 
of the numerical relations of the parts as a basis for classification 
in separating from them the Edwardsias and referring the thus 
restricted groups to the Actiiiinje. 

Andres enters into a iimcb more minute division of his families into 
8iibfaini】ies, many of wbicb are well founded, but it will not be con- 
venient to criticise them here. 

With Aucjr^i^ the second period in the history of the classification 
of the Anthozoa may be said to close. The period was marked by a 
gradnally growing tendency to divide the group into a number of 







Proc, 1^, M, 93- ~« 9 

Digitized by 


equivalent subgroups, and, so far as the Actinians are concerned, by 
the increase in the number of recognized genera and their division into 
families, subfamilies, etc. The distinguishing characters of the various 
groups were drawn for the most part from external characters; impor- 
tance, for instance, being placed upon the presence or absence of a 
corallum, whether the base is adherent or not, iu the shape of the ten- 
tacles, etc. Comparatively little was done toward attaining a thorough 
knowledge of the anatomical relationships of the various parts, or per- 
haps it would be better to put it iu this way, that the anatomical 
knowledge 力 hat had been acquired was not sufficiently extensive to be 
employed for systematic purposes. The names of Hollard, Quatrefages, 
Haime, Thorell, Teale, Schneider and Eotteken, Stoliczka, etc., recall 
important additions to our knowledge of Actinian morphology, but the 
observations were not sufficiently extended to have suererested. the im- 
portance that should have been attached to them. 

We are now iu the third period, so brilliantly introduced by the 
brothers Hertwig with their mouograph on the Actinians ('79). The 
period in its beginning overlaps, consequently, the second period. The 
fundamental characteristic of Actinian classification at present is the 
foundation which it possesses on anatomical and phylogenetic features. 
The arrangement of the mesenteries and their ontogenetic succession 
are the criteria which serve to separate the larger groups, and these 
criteria have been extended, so far as our present knowledge allows of 
it, to the group Anthozoa as a whole. The first step iu this direction, 
as stated, was made by the Hertwigs (,79), who, as a result of their 
observations on a number of Actiuiaria, arrived at the following con- 
clusion : 

Bei der Eintheilung der Anthozoen sind die Septen in erster Reihe zu beriicksich- 
tigen, aber weniger die Zaiil als vielmehr der Bau, die Anordnung derselben um das 
Schlundrohr und ihre Entwicklung. Wenn wir von dieser Gnindlage ausgehen, 
werden die Anthozoen voraussichtlich in iiiehr als 2 Ordnungen zu zerfallen sein. 
Mit Erfolg aber kann ein neues System erst dann aufgesiellt werden, wenn die ver- 
schiedenen Familieu der Zoantharien, der Fleischpolypen sowohl als der Korallen, 
auf die Morphologie ihrer Septen, iiber die wir vielfach noch gar nichts wiseen, nach 
alleu Richtungen untersucht sein werden. 

The Hertwigs recognize five groups of Anthozoa, based on the charac- 
ters indicated in the above quotation, viz, the Actinidae, Edwardsiae, 
Zoanthidae, Cerianthidae, and Alcyonaria. As regards the Madrepo- 
raria, they do not commit themselves definitely, recognizing the paucity 
of the information with regard to their anatomical peculiarities at their 
disposal; at the same time, however, they consider it probable that 
when the required information is acquired the group of the Zoantharia 
sclerodemata will be split up, a large part of the corals being associated 
with the Actinidse, others, perhaps, with the Zoaiithinfe, and others 
with Edwardsiae, while others, again, may show an arrangement and 
structure of the mesenteries peculiar to themselves. 

VOL. XVI, 1 
薦. J 



The idea brought forward in this work was elaborated more folly by 
Richard Hertwig in his report on the Challenger Actiniaria ('82), in 
which the structural peculiarities of the various forms are employed, 
not only to distinguish the principal groups, but also to define in an 
accurate maimer the various families of the Hexactinife. In some par- 
ticulars the idea was carried a little too far, owing to the absence at 
that time of anatomical studies of a large series of forms, Hertwig be- 
ing obliged to rely entirely on his own observations in deciding as to 
the relative importance of a character. Omitting the Alcyonaria and 
Antipatharia from consideration, Hertwig recognizes six tribes of Acti- 
uiaria, which correspond in taxonoiiiic value to Andres's families. These 
tribes are (1) HexactiniaB, (2) ParactinijE, (3) Mouauleae, (4) Edwardsiae, 
(5) Zoanthese, and (6) Ceriautheae, and all are characterized by the ar- 
rangement of the mesenteries. Three of these orders correspond to 
families of Andres' classification; other three of Andres' families, viz; 
Thalassianthinse, StichodactyliiiSB, and Minyadiute are grouped with 
his Actininse to form the tribe Hexactini*. while two other tribes, not 
represented in Andres' system, are instituted for forms presenting an 
arrangement of the mesenteries not previously recognized. In com- 
paring the systems of Hertwig and Andres, however, it must be re- 
membered that the two works were so nearly contemporaneous that 
the respective systems were entirely independent one of the other. 
Andres, it is true, had the advantage of the earlier work of the 
brothers Hertwig (,79), which no doubt influenced considerably liis 
ideas as to the relationships of certain of the groups, but had uo cog- 
nizance of Richard Hertwig's later observations. 

The introduction into the classification of the Anthozoa of a system 
based upon anatomical peculiarities, instead of one resting entirely on 
variable characters, readily subject to modification in accordance with 
external conditions, was very important. There yet remained to be 
taken the farther step of adding to anatomical characters the informa- 
tion derived from embryological investigation, a step the importance 
of which the Hertwigs had recoguized and contributed to, to a certain 
extent. Some of the necessary information was contributed later by 
Boveri ('90) and myself ('91), and as the result of these observations 
I drew up a classification of the Anthozoa founded upon structural and 
embryological characteristics. Either of these classes of facts, taken by 
itseli, IS liable to lead to errors; it is only by conibiuiug both that a 
true knowledge of the phylogenetic relationships of the various groups 
can be obtained. For instance, relying entirely on embryological data, 
the Hexactiiiife could bo separated into three distinct groups, one in- 
cluding those forms in wliich the mesenteries appear according to the 
succession described by Lacaze-Dnthiers ; a second, in which the 
mesentery succession is that described by the Hertwigs (79); and a 
third, in which it is that described by Haddon ('87), H. V. Wilson 
('88), and myself ('91). I have shown, however ('91 a), that the third 


method is to be regarded as the typical oiie, and that the orders of 
succession described by Lacaze-Diithiers and the Hertwigs are secon- 
dary modifications of this, called forth by peculiar conditions; and, 
furthermore, anatomical investigation of forms developing in these 
various manners shows so much similarity iu them all as to do away 
with any idea of classing them in three distinct groups. 

The classification which I proposed differs from that of Hei twig in two 
particulars. In the first place I disregard his tribe Paractiniae, which I 
have shown to be uunatural and untenable, and I gioup the form upon 
which his tribe Monaiileie was founded with the Gonacti/nia, long before 
described by Sars and later studied iiiore thoroughly by Blochmann 
and Hilger ('88), and with the Oraetis Diomedew, described in subse- 
quent pages of this report, forming thus a tribe, the Protactiuiae, the 
members of which I take to represent stages in the phylogeny of the 
HexactinisB. Strictly speaking, perhaps each of these three forms should 
constitute an order, but is seems to contribute to the convenience of the 
classification, without introducing any confusion, to group them together. 
I recognize the following tribes of Antbozoa : 

The propriety of considering the Eugosas as forming a tribe equiva- 
lent to the Alcyonaria, for instance, is open to question, since we nat- 
urally know nothing: as to their soft parts and can only form an ex- 
ceedingly uncertain idea of Low they were arranged from the arrange- 
ment of the septa in the coralliim. The Antipatharia form a natural 
group, apparently, though it is uncertain what their affinities with the 
other groups may be. The remaining tribes seem to have their phylo- 
geiietic relationships fairly clearly defined.* 

*In a recent paper Ed. van Beneden ('91) contests the idea that there is a phylo- 
genetic connection between the Ceriauthete and the Edwardsise and Hexactiniae. 
On pages 140-141 of his paper he sums up the differences which the CeriantheaB show 
to these forms, and it may not be out of place here to consider the value of these 
supposed differences. Difference No. 1 does not require consideration, since it 
stands or falls with the accuracy or erroneousness of No. 2. This is as follows : If 
the snlcar directives are designated as I and the remaining mesenteries of a twelve- 
mesenteried Hexactinian are designated according to their succession, counting from 
the siilcar directives towards the sulcular, as II, III, IV, V, and VI, then the em- 
bryonic succession of the mesenteries iu the Hexactinije is III, V, I, VI, II, IV, wJiile 
ill the CeriantheaB the succession of the first twelve mesenteries is II, III, I, IV, V, 
VI. The fallacy of this is evident'. It has not been claimed that the first twelve mes- 
enteries of Ceriantheie and Hexactiniae are homologous, but only that the first eight 
in both groups 'are homologous with the eight Edwardsia mesenteries. Considering 
the embryonic succession of these mesenteries in both groups, it will be found to be 
identical, thus: III, I, II, IV. Difference No. 3 refers to the presence of longitudinal 
(adductor) muscles on the mesenteries of the Edwardeiao and Hexactinit© and their 
absence on tboee of tli^ Ceriantheie, and to the presence of ectodermal longitudinal 

1. Riigosap. 

2. Antipatharia. 

3. Alcyonaria. 

4. Edwardsift. 

5. CeriantheaB. 

6. Zoaiithejp. 

7. ProtactiniaB. 

8. HexactinisB. 

TOL. XVI, "1 
1893. J 



A detailed criticism of the various families which have been proposed 
is not necessary, since this will be entered into in the descriptive por- 
tion of the report, so far as certain families of the Hexactiuise are con- 
cerned. Andres ('83) added a considerable number of families to those 
which had previously been recognized, and the majority will, no 
doubt, stand. Hertwig too has added a number of new families, and 
at the same time has given aii interesting criticism of Andres' classifi- 
cation aud a comparison of it with his own. Many of the families 
Andres recognized are more accurately defined, aud atteution Jias been 
called to the criteria upon which families should be based. One of 
the most recent classifications, is that of Danielssen ('90), which is 
essentially that of Ilertwig ('82), confused, and without the corrections, 
which Hertwig ('88) subsequently introduced. In fact, it must be 
acknowledged that Danielssen's work is a ^?ieat disappointment, in 
that the descriptions are given in such a manner as to preclude con- 
fidence in their accuracy, while the figures illustrating them are beau- 
tiful examples of "how not to do it." The tribe ^Egireae, which Dan- 
ielssen proposes, certainly requires farther study before being accepted, 
and the same remark applies to his families Sideractidae, Madouiac- 
tidse, and Andvakidae. 

I shall content myself with stating the families which I believe to 
be worthy of recognition, making some brief remarks on their limita- 
tions, and on certain somewhat doubtful forms. 

I think it convenient to consider the mode of arrangement of the ten- 
tacles of classificatory importance, and to recognize two subtribes of the 
Hexactiniai to which Andres' uaiue may be applied : Actiiiinae to those 
to which the tentacles are arranged in cycles, and Sticliodactylinae to 
those in which they are arranged radially. To these two I added ('89) 
a third, the Dendromeliua?, which is hardly of equal value, and which it 
will be better to reduce to the rank of a family. It includes forms which 
possess dendritic or g^lobular processes or arms projecting from the 
coluipn wall below the margin, such as are found, for instance, in Le- 
brunea, OphiodiseuSj and Viatrlr* 

The I would not, however, adopt as a subtribe, since 
they differ from the Actiuina^ only in the compound character of their 
tentacles, and a passage to them is furnished by the member of Andres, 
subfamily Heteractidfe. This family is, however, not altogether natu- 
ral since the genus Ragactis must be removed from it and referred to the 

muscles in the latter and their absence in the other two groups. The absence of 
" adductor " muscles in the CeriantbeaB is a qnestion of observation, since Boveri 
has described and figured them; and with regard to the presence of ectodermal 
longitudinal muscles in the Hexactiniop, van Beneden has apparently overlooked 
Hertwig's account ('88) of their occurrence in Corynactia sp? and Corallimorphm 
ohtecius. In the present report I describe tbeir occurrence in Halcurias piJatus. 

*It seems fairly certain that the actiiiian recently described by H. V. Wilson 
ぐ 90) as Hoplophoria coraU'tgens (sic) is identical with the Viairix glohuUfera originally 
described by Ducliassaing and Michelotti ('60). 


Sagartidae, as was demonstrated to me by the late Dr. J. I. Nortlirup, 
who discovered the Sagartian character of R. lucid れ, an observation I 
have since been able to confirm.* The genus Uloactis also proposed 
by Andres ('83) for the llyanthus mazeli of Jourdan (,80) seems to belong 
rather to the Halcampidae, E. Mazeli being apparently nearly related to 
H, producta of Stiinsoii. 
The following is the classification I suggest : 


in cycles. 



Sphincter Jib- 
sent or weak 


not numerous . . HalcaiupidaB. 

numerous AntheadaB. 

Sphincter endodermal ; 
tentacles decid uous Boloceridae. 

Sphincter iiieso- ' 
gloeal ; 

Sphincter en do- ( 
dermal c ir- < 
ciimscribed. < 

No acoutia Paractidae. 

Acontia Sagartidse. 

Acmrhagi wart-like. BuuodidaB. 

Acrorhagi foliate. Phyllactidae. 

( Tentacles simple .Heteractidte. 
Tentacles warty or branched \ Tentacles 

( compound . Thalassiauthid sb . 

Tentacles reduced to stomidia \ PolyopMae. 

infill provided in its upper part with branched 

or globular processes Deudromelidae. 

Free-swimming forms MinyadaB. 

r Tentacles few, capitate Corallimorphidte- 

' Tentacles all of J Tentacles numerous, cylin- 

one form. ) drical DiscosomidsB. 

[Tentacles nodulated Aurelianidfe. 



Tentacles a: 
ranged radi-^ 

Stichodactylinae . 

Tentacles of two 

Marginal tentacles cylin- 
drical; disc tentacles 
wart-like, branched, or 
foliate Rhodactidae. 

Marginal tentacles pinnate, 
disc tentacles "wart-like - - Phy manthidse. 

^ Tentacles of various forms, not cylindrical. - . CriptodeudridsB. 

丄 have chosen the term Halcampidae in preference to that of Ilyan- 
thiaae because we are at present ignorant of the anatomical charac- 
teristics of Ilyanthm ) it will, however, probably prove to be similar to 
Halcampa in many respects, in which case the older term should be re- 
stored. The Siphonactinidse, I think, should be fused with the Hal- 
campidae, the presence or absence of a conchula not being of sufficient 
moment for family distinction. 

The Boloceridae is a new family, for whose existence reasons will be 
given in Part IL The Antheomorphidtne of Hertwig ('82) I include 

* It seems doubtful, however, if the /?. pulchra of Andres is likewise a Sagartid. 
I should rather be incliued to consider it a Phymanthtis, since it presents certain 
striking resemblances, judging from Andres' description, to P. crucifer. 

Digitized by 




for the present with the Antlieadae. My reasons for placing the Phyllac- 
tidae among the ActininaB have been given in another place ('89a). 
Hertwig's Polyopidte and Sicyonidae I have placed in the Actininae, 
not recognizing his tribe, Paractiniae. These forms require further 
study. The Lipouemidae are too much open to suspicion to be accepted, 
as will be seen from what is said in Part ii concerning Bolocera hre- 
vieorniSj Idponema recalling strongly a Bolocera^ while Polysiphonia re- 
calls Actinernus, 

Under the Sagartidae is included the Phellidse, which may be re- 
garded as a subfamily under Haddou's name of Ghondractminae, and 
the Amphianthidae, which are probably all referable to the Sagartidae 
and to the subfamily Ohondractiuinae. 

The MinyadsB are inserted with the ActininaB as a family, but little 
is as yet knowu of their anatomical peculiarities. 

The classification of the Stich odact y lin ae hardly calls for comment, 
except to point out that, of the Aurelianidae and Criptodendridae very 
little is known, nothing indeed as to anatomical characters. I have 
employed the form of the tentacles, following Andres with modifica- 
tions, as a basis for the classification; but in those groups members of 
which have been studied it has been found that more or less definite 
anatomical features are associated with the various tentacular modiii- 

This classification is, it must be understood, intended to be purely 
tentative and to take coguizanca only of families which seem well 
authenticated. doubt the changes and additions which will be 
required to make it at all accurate are numerous ~ how numerous 
future observation will determine. 

Actinozoa not forming colonies ; with eight mesenteries, three of which 
on each side have their longitudinal muscles upon their snlcar faces, 
while the other two, situated at the sulcar surface, have these muscles 
on their sulcular faces. Tentacles simple, usually more numerous than 
the mesenteries. 

I do not consider it necessary at present to divide the Edwardsias 
which we know Into two genera, as Andres ('83) has done, much less 
to make the number of the tentacles the feature upon which to base 
such a division, since this is a character liable, to judge from the de- 
scriptions of species which we possess, to numerous gradations. When 
a thorough anatomical study has been made of a number of different 

Part II. 


Tribe EDWARDSIiE, Hertwig. 

Genus EDWARDSIA, Quatrefages. 
With the characters of the tribe. 


species, it may be found advantageous to make a division; at present 
it does not seem advisable. - 

No. 704. station 2783. Lat. 51。 02' 30" S. ; long. 74° 08' 30" W. Depth 122 fathoms. 
1 specimen. 

The single specimen for which I propose the above name was strongly 
contracted, the entire capituluin being introverted. In this contracted 
condition (PL xix, Fig. 1.) it measured 1.7。*" in height, and its greatest 
diameter is 0.45。《». 

The physa is rounded and translucent, allowing the mesenteries to 
show through. The scapns is covered by a thin, brown, chitin-like 
"epidermis," resembling, apparently, that covering U. Claparediij but 
unlike it, being almost smooth. It is quite translucent and consists of 
two layers (PI. xix, Fig. 4) ; on the outside is a thin layer of foreign matter 
(/), consisting of very fine sand particles, spicules, etc" and below this 
a cuticle-like layer (eu) covering the ectoderm (ec) and sending here and 
there into that layer prolongations which seemed occasionally to unite 
with the mesoglcea. The arrangement is very similar indeed to what I 
have described for Zoanthus sociatm ('89), though it is not so certain 
in this case that the cuticle is really a portion of the mesogloBa. The 
ectoderm (PL xix, Fig. 4, ec) consists of cells, not at all columnar, as is 
usually the case in the Actinozoa, and shows no trace of either gland 
cells or nematocysts. 

The scapns is marked by eight longitudinal furrows, corresponding 
to the insertions of the mesenteries, and the intervals between these 
furrows are occupied by numerous irregularly scattered clear spots, 
which recall the tubercles described by Andres ('80) and Danielssen 
('90). Their structure is, however, somewhat different from what these 
authors have described. The ectoderm over a small area is slightly 
thickened and projects through the covering investment, but no nemato- 
cysts were to be found in it. The transparent appearance which is so 
characteristic of the tubercles is due to a comparatively large oval cavity 
in the mesogloea, lying below the tubercle and always separated, 
apparently, by a very delicate layer from the ectoderm, though a small 
collection of granules and, in some cases, a few cells are to be found in 
the cavity. (PI. xix, Fig. 4). 

In transverse sections it is seen that a portion of the scapus is intro- 
verted as well as the entire capituluin; sections taken at a little more 
than 3随 from the upper extremity of the contracted animal show the 
cuticular investment which is characteristic of tbe scapus. In this in- 
troverted region, however, the layer of foreign material (PL xix, Fig. 2,/) 
is very much thicker than on the outer surface of the body, and, further- 
more, in each interval between the insertions of two mesenteries a 

1. Edwardsia intermedia, sp. nov. 

Plate XIX, Figs. 1-4. 

TOL. XVI, 1 
1893. J 



strong ridge, formed principally of mesogloea, projects, and as the 
capitulnm is approached cavities appear in the ridge, giving it in cross- 
section a club-shaped outline. 

The capitulnm is apparently very short and is destitute of any in- 
vestment. JTlie number of the tentacles I could not ascertain, but they 
seem to be few, perhaps eight, almost certainly not more than sixteen. 
Tbey project down into the stomatodaeum in the manner described by 
Quatrefages ('41). 

The storaatodaeum is short and is slung by the eight mesenteries, whose 
musculature has the usual arrangement. All the mesenteries are gono- 
phoric and possess mesenterial filaments. Since Andres ('80) has stated 
that in Claparedii the respiratory portions of the filament are wanting, 
I may state that in the species here described they are unmistakeably 
present, though short. The bases of the mesenteries at their insertion 
into the column wall are furnished with pinnately arranged muscle proc- 
esses (PI. XIX, Fig. 3 h m). The longitudinal muscles are strong, resem- 
bling in transverse section those of E. tecta as figured by Haddou ('89). 

It is impossible to identify this form with any of the species that 
have been described. Within recent years a number of Edwardsias 
from deep water have been described by Moseley ('77), Marion ('82), 
E. Hertwig (,88), and Dauielssen ('90), but the descriptions are not in 
all cases sufficiently detailed to permit of a correct idea of the mor- 
phological characteristics. The structural features which are of im- 
portance for classificatory purposes seem to be the tubercles, tlie shape 
of the longitudinal and basal muscles of the mesenteries, the presence 
or absence of longitudinal ridges on the column, and, what is probably 
of less importance, the number and arrangement of the tentacles. 

E. intermedia agrees, as already stated, with E, tecta (Haddon, ,89) 
in the structure of the longitudinal muscles, but differs from it in pos- 
sessing tubercles and in the shape of the basal muscles of the mesen- 
teries 5 it approximates E. fusca Danielssen ('90) in the number of the 
tubercles, though they are not arranged with anything like the regu- 
larity which they have in Daiiielssen's figure, and in addition the shape 
of the longitudinal muscles is altogether different ; it resembles E, car- 
nea (Haddon, ,89) in tlie possession of longitudinal ridges on the 
capitulum and upper part of the scapus, but differs from it altogether 
in the shape of both longitudinal and basal muscles. 

In consequence of its possessing certain of the characteristics of each 
of these three species I have named the form here described E. inter- 

Aiitliozoa with twelve primary mesenteries, of which eight at least 
are perfect, and which are arranged in pairs, the longitudinal mesenter- 
ies of each pair being on tlie faces of tlie mesenteries which are turned 



towards the intramesenterial space, except in the case of two pairs, 
the directives, situated at the extremities of the sagittal axis of the 
stoinatodaenin, whose longitudinal muscles are on the faces of the me- 
senteries whicli look towards the adjacent intermesenterial space. 

also present 5 of these there may be oue on each side, situated in the 
sulculo-lateral iutermesenterial space, or a pair on each side in the 
same iutermesenterial space, or two pairs on each side in the sulculo- 
lateral and lateral iutermesenterial spaces. The development of the 
mesenteries is upon a bilateral plan. 

Protactiniae with twenty mesenteries, twelve of which are primary, 
and two pairs on each side, in the sulculo-lateral and lateral iuterme- 
senterial chambers respectively, secondary. Only the eight primary 
mesenteries corresponding to the Edivardsia meseuteries, are perfect, 
gonophoric, and provided with mesenterial filaments. 

No. 727. Station 2839. Lat. 33^ 08' N. ; long. 118^ 40' W. Depth, 414 fathoms. 
Several specimens. 

All tlie specimens are contracted extensively (PI. xix, Figs. 5 and 6), 
and measure in this condition 5 to 8""" in height and 10 to 13*"™ in 
diameter. The base and column are colorless and translucent, allow- 
ing the internal organs to show through, but sections show that the 
disc and tentacles have yellow pigment granules in their ectoderm, and 
probably in the living condition they bad a more or less decided yellow 

The base is more or less rounded (PL xix, Fig. 5) and passes directly 
into the column, there being* uo limbiis. The column is marked by 
twenty longitudinal grooves corresponding to the insertions of the 
mesenteries. At the summit of the contracted animal ten tubercle-Uke 
processes can be seen surrounding the entrance into the cavity con- 
taining the contracted tentacles, and in sections these tubercles may 
be seen (PL xx, Fig. 11, tti) to be due to thickenings of the mesogloea. 
Ill some of the specimens they appear to be infolded along with the 
tentacles. The ectoderm of the cohinin wall has entirely disappeared. 
The inesogloea is thin, and more or less nbrillar in structure witli com- 
paratively few cells. The sphincter muscle is of the diffuse type (PI. 
XX, Fig. 11.), its mesogloeal processes being long and numerous, so 
that a fairly strong muscle is produced. 

The tentacles appear to be ten in mrniber. They are rather short, 
cyliudrical, obtuse. In transverse sections it is seen that their longi- 
tudinal muscles are confined to the ectoderm, and for the most part are 

Genus ORACTIS, gen. nov. 

2. Oractis Diomedese, sp. nov. 

Plate XIX, Figs. 5-8; Plate xx, Figs. 9-11. 



of moderate strength, but towards the base two regions are to be found 
upon the outer surface of the tentacle where the muscle processes reach 
an excessive development (PI. xix, Fig. 7.) In sections which pass 
through the point of origin of the tentacles, just where they arise 
from the disc (PI. xx, Fig. 10), it can be seen that these two muscle 
bundles (m) are continued upon the disc, forming strong niuscies lying 
immediately over the mesenteries, one bundle of each tentacle coming 
irom the mesogloea over each of the mesenteries which limit the in- 
tramesenterial space to which the tentacle belongs. These muscle bun- 
dles are not, however, continued to any extent upon the disc towards 
the mouth opening, but appear to be confined to the peripheral region 
where the tentacles arise. 

The stomatodseum (PL xix, Fig. St) is rather short, arid has only one 
siphonoglyphe which is deep, its mesogloea being much thicker than it 
is elsewhere on the stomatodeuui. The remainder of that structure is 
marked by six longitudinal ridges, each of which corresponds to the 
insertion of a mesentery. 

As is indicated by the furrows of the exterior of the column there 
are twenty mesenteries. Eight of tbein are perfect, goiiophoric, and 
provided with mesenterial filaments, while the remaining twelve are im- 
perfect, sterile, and destitute of filaments. The arrangement of the 
mesenteries is exceedingly interesting (PI. xix, Fig. 8). There are two 
pairs of directives, having the characteristic arrangement of the 】0Dgi- 
tudinal muscles; that pair (III) which is attached to the siphonoglyphe 
marks the sulcar surface of the body. On each side of the sulcar 
directives is au imperfect mesentery (V) with its longitudinal muscle 
upon its sulcular surface, and succeeding this comes a perfect mesen- 
tery (I) forming with the imperfect one a pair. Then follow a pair of 
imperfect mesenteries (VIII), then a pair formed by a sulcar imperfect 
(VI) and a sulcular perfect mesentery (II), then a pair of imperfect mes- 
enteries (VII), and finally the sulcular directives. 

It must be stated that the figure I have given is to a certain extent 
diagrammatic, inasmuch as in a section through the stomatodseum the 
longitudinal muscles of the imperfect mesenteries could not be readily 
made out, while further up the column, in sections which passed through 
the column and disc, they were well developed. I have represented 
therefore the arrangement as regards the perfectness or imperfectness 
of the mesenteries as seen in a section passing through the stonia- 
todaeum, but the musculature as seen in sections passing through the 
column and disc. 

The loiigitudinal muscles are not strong, and in the perfect mesen- 
teries occupy the greater part of the surface (PI. xx, Fig. 9). The 
endoderm of the mesenteries presents a rather peculiar vacuolated ap- 
pearance, reminding one of the structure which it presents in Cerianthtis, 
The mesenterial filaments which are developed only on the eight per- 
fect mesenteries appear to lack the " Flimmerstreifeii ,, but T can uot be 


certain that they are really absent. The contraction of the specimens 
renders it difficult to understand the exact structure of tbe filaments. 
The ova are large and contain a considerable amount of food-yolk. 

The significance of the arrangement of the mesenteries of this form 
I have elsewhere pointed out ('91). The eight perfect mesenteries 
evidently correspond to the eight Edwardsia mesenteries; the imper- 
fect mesenteries which form pairs with adjacent perfect ones (I and II), 
are evidently the mesenteries which convert the octameral into the 
dodecameral condition with paired mesenteries. The imperfect pairs 
yil and VIII are secondary mesenteries and arise in pairs in the two 
intermesenterial chambers nearest the sulcular directives. 

If we omit pair viii we have the arrangement which occurs in Oonac- 
tinia (Blochnfann and Hilger, '88), and if the sulcular member of 
pair VII be omitted the condition obtaining in Scytovhortis ( Bertwig, 
,82) will appear. It seems that these two forms, together with Oract", 
represent three links in the chain by which the ancestor with twelve 
mesenteries, all of which arose singly aud bilaterally, becomes con- 
verted into the Hexactiniae, in which the muscles arise in pairs and 
radially. In Scytophorus tbe original method of formation is carried 
over into the formation of the single secondary mesentery. In Oonac- 
tinia the paired mode of formation is beginning, in Oractis it is 
thoroughly established, but in both these forms the bilateral mode still 
holds. Finally, in Halcampa the mesenteries arise in pairs and radially. 

It will not be necessary to enter into the details of this idea here, 
since it lias been treated of elsewhere in connection with some other 
facts ('91a). It may be well, however, to point out that there is embryo- 
logical evidence to show that the secondary mesenteries of the Hexac- 
tiniae make their appearance in the sulculo-lateral chambers earlier 
than in the others, and those of the lateral chambers develop before 
those of the sulco-lateral ones (Dixon, ,89), a succession which exactly 
corresponds with the phylogenetic development seen in the Protac- 

Actinozoa with six, eight, or ten pairs of perfect mesenteries, which 
form a principal cycle, and to which may be added a varying number 
of additional cycles, perfect or imperfect, the mesenteries of which 
develop in pairs and radially, appearing almost simultaneously in all 
the intermesenterial spaces. The longitudinal muscles of each pair 
are on the faces which look towards the intramesenterial spaces, except 
in the case of two (occasionally one) pairs, the directives, which are 
attached to the two (occasionally one) siphonoglyphes, and whose longi- 
tudinal muscles are on the faces which are turned towards the adjacent 
intermesenterial spaces. 

The above definition differs considerably from that given by Hertwig 
('82), who founded the tribe. This results from the fact that I include 


1893. J 



within it not only hexamerous forms, but also octamerous and decam- 
erous Actiiiians. lu other words, I fuse with the original Hexactiniae 
Hertwig's tribe Paractiniae, which is altogether artificial and unneces- 
sary. My reasons for this opinion have been given at length elsewhere 
and need not be repeated ('89, ,91). 

Hexactiniae in which the tentacles are arranged in cycles, oiily a sin- 
gle tentacle communicating with each endocoel. 

Actininae with a small number oJt mesenteries, six, ten, or twelve pairs 
being all that are present; longitudinal muscle peimous narrow, but 
strong; no special sphincter muscle ; eoncbula presenter absent; base 
usually rounded and vesicular. 

In his monograph, Andres ('83) divided the family Jlyauthidae, which 
bad been previously limited by the exclusion of the Ceriantheje and 
EdwardsiaB, into three distinct families, or, as he termed them, sub- 
families. One of these, the HiilciMiipidiB, contained the genus Halcampa^ 
the IlyanthidsB included only the genus Ilyanthm^ while for those forms 
which possessed a conchula the family Siphonac^tinidae was established. 

A more recent author, Haddon (,(S9), seems to regard this last fam- 
ily with uncertainty. At all events he removes from it and associates 
with the HalcampidiB the genus Feachia, recognizing, however, the 
the possible necessity for the establishment of a separate family for it. 

The uncertainties which interfere with the delimitation of the family 
Halcainpidje are mainly two. Are forms which possess a conchula to 
be associated with others which do not have this structure, but whose 
mesenterial arrangement is similar? And are decameroiis to be asso- 
ciated with hexamerous forms ? I believe that both these questions 
should be answered affirmatively. The forms belonging to the family 
Siplionactinidae, so far as they are known, agree in certain important 
features, viz, in the usual absence of an adherent base, the absence of 
a sphincter, the small number of mesenteries, and the strong though 
narrow longitudinal muscles, with the members of the genus Halcampa, 
and differ from them oii】y in the possession of a conchula, a structure 
of probably comparatively little morphological importance. As regards 
the arrangement of the mesenteries, even if we separate the forms with 
a conchula from those without it, it will be necessary to associate 
together hexamerous and decamerous species, unless we wish to multi- 
ply families beyond convenience and necessity. Halcampa endromitata^ 
etc., arehexamerous, and H.producta is decamerous among the Halcamp- 
idae; and among conchula-beariii g forms Peachia hastata is decamerous, 
while Bictdium parasiticum is hexamerous, possessing twelve pairs of 

Order ACTININ^, Andres. 



I think that the purposes of classification will be better served by 
uniting these and similar forms into a single family, for which the name 
Halcampidae, already used iu this sense by Haddon ('89), may be 
employed, and to recognize in this family several genera. The genus 
Halcampa seems to be well characterized by its hexamerism and the 
distinction into capitulum, scapus, and physa. H. producta, of the east 
coast of North America, and' H. capensis Verr., H. brevicornis (Stimp- 
son), and R. Stimpsonii Verrill (,65), decamerous forms, may be 
referred to another genus, though probably it will be well to separate 
H. producta from the other three and refer it to a separate genus. The 
form described by Jourdan ('80) as Ilyanthm Mazeli seems to be 
closely related, and H, producta may be referred with it to the genus 
Eloactis proposed by Andres ('83). 

Iu addition to these three genera, since Andres' genus Halcampella 
and Danielssen's Halcampoides ('90) do not seem necessary, there will 
be of the conchula-bearing forms Peachia, which is decamerous, and 
Bicidium, which is liexamerous. The genus ActinopsiSj which is asso 
ciated with these two genera by Andres, presents certain external 
characters which make one hesitate somewhat to retain it in the group. 
Until an anatomical study of it has been made it will probably be as 
well to leave it where it is. 

Among the Albatross material I find two species which may be re- 
ferred to the family thus emended. Que of these is a Peachia; the 
other must, I think, be referred to a new genus related to Halcampa 
or Eloactis, 

' Genus HAIiCURIAS, gen. nov. 

Halcainpidfe with an adherent base; column cylindrical; tentacles 
numerous and short; ten pairs of mesenteries, all of which are perfect, 
though four pairs situated iu the sulco-lateral and lateral intermesen- 
terial spaces are less extensively developed than the other six. No con- 

3, Halcurias pilatus, sp. iiov. 

Plate XX, Figs. 12-13. Plato xxi, Figs. 14-15. 

No8. 708, 709. Station, 2785. Lat" 48" 09' S. ; loug., 74" 36' W. Depth, 449 fathoms. 
Three specimens. 

The base is flat and adherent, one of the specimens being seated 
upon a piece of dead coral. The column (PI. xx, Fig. 12) is cylindrical, 
much wrinkled from contraction, but apparently possessing no warts 
or tubercles. It measures 2.3细 in height, while its diameter at the 
base is 2。™. Slight indications of longitudinal bands of color can be 
perceived, but they are exceedingly obscure and could not be discovered 
on all the specimens. 

The margin is smooth and forms a very distinct parapet around the 
bases of the tentacles. These are numerous, amounting to about 
seventy in one specimen in which they were counted, and are arranged 

Digitized by 



in about three cycles. They are simple, cylindrical and taper to a point, 
and cover almost the entire disk. There being no special sphincter 
muscle, the tentacles are not covered in contraction. 

There is no conchula, and only one siphouoglyphe, which is neither 
very deep nor well defined. The surface of the stomatodaeum possesses 
numerous ridges, which are high (PI. xxi, Fig, 14, st) and may bifurcate 
at the extremity or give off secondary ridges. They are more numer- 
ous than the mesenteries, and do not seem to stand in any very 
definite relation to them. The mesenteries are twenty in number. 
They are arranged in pairs, two of the pairs being directives, and are 
all perfect. Below, however, it is seen that four of the pairs, as in 
PeacMd, are much narrower than the other six, these narrow pairs being 
situated in the sulco-lateral and lateral intermesenterial spaces. The 
mesenteries are thin; at the base there are pinnately arranged muscle 
processes (PI. xx, Fig. 13, ftm.), and separated from these by a region in 
which the mesentery is exceedingly thin are the longitudinal muscles. 
These are very strong (PL xx. Fig. 13), but at the same time narrow, 
forming a strong protuberance upon the surface of the mesentery. 
Above, however, they widen out (PI. xxi, Fig. 14) and the processes are 
not so high. 

All the mesenteries bear reproductive organs. 

There are a few points in the histology of this species which are inter- 
esting. The inesogloea is fibrillar, especially towards its inner surface, 
and contains very numerous cells. It is in the ectoderm however, that 
the most interesting peculiarities appear. The ectoderm of the column 
wall is high and contains, as usual, many gland cells. In addition to the 
usual elements, however, it also contains numerous nematocysts (PI. xxi, 
Fig. 15, n) lying iu its outer portion, sometimes very closely crowded 
together. Immediately external to and resting upon the mesogloea, 
roundish bodies— or, rather, bodies appearing round iu cross-section 
(mf.) 一 which stain somewbat deeply, can be perceived. These seem to 
be muscle fibres, having a longitudinal direction. They have all the 
appearance of muscle fibres, but I was not able to render their nature 
certain by the study of maceration preparations. Father evidence for 
their muscular nature is, however, to be found in the presence, exte- 
rior to them, of a thin layer of fibrillae having all the appearance of a 
nerve layer. 

Longitudinal muscles and a nerve layer are, as a rule, absent in the col- 
umn wall of the Hexactinise ; but, ou the other hand, are well developed 
in the Ceriantheae, and it seems probable that the more primitive Actino- 
zoa likewise possessed them. Hitherto they have been found among 
the Hexactiniae only in Corynactis f sp? and Gorallimorphus obtectus, in 
which forms they have been described by Hertwig ('88). The fibres 
of Halcurias resemble those of Gorallimorphus in being poorly developed, 
and are apparently fewer in number. In Corynactis f on the other 
Imnd, they seem to reach a fair degree of development. 


A few words are necessary regarding the affinities of this form. It 
differs from all other genera of the Halcampidae by its adherent base 
and by the large number of tentacles which it possesses. Actinopsis 
possesses the same characteristics, although the tentacles are much 
longer in proportion, but differs in haviuff a conchula. There reason 
to doubt, however, whether Actinopsis can be referred to this family. 
Among the members of the family, however, indications of an adherent 
base are found, as in Eloactis prodncta, and the importance of this char- 
acter seems to be far outweighed by the small number of the mesen- 
teries and the structure of their muscles. It seems tolerably certain 
that the Halcampids are the simplest and probably the most primitive 
of the Hexactiniae, and the presence of longitudinal muscle fibers in 
the ectoderm of the column wall of Halcurias is a primitive character- 
istic. I think, on the whole, that it is to be regarded as much more 
nearly related to the Halcampids than to any other family of Hex- 

Genus PEACHIA, Gosse. 

HalcampidaB, with rather short tentacles, few in number; with four 
pairs of narrow sterile mesenteries, situated in the lateral and sulco- 
lateral intermesenterial spaces, and six pairs of perfect fertile mesen- 
teries; and with a single deep siphonoglyphe. Longitudinal muscles 
of the mesenteries strong. Conchula present. 

Go8se ('55) instituted this genus for the reception of P. hastata and 
Halcampa chrysanthellum^ later on, however ('58), removing the latter 
form to the genus to which it is uow universally assigned. Andres 
(,83) employs, instead of Gosse's name, that proposed by Koren and 
Danielssen ('56), Siphonactiniaj but the term proposed by Gosse has un- 
doubtedly the priority, as Haddon points out ('84). In his revision of 
the British Actiniae, Haddon ('89) gives a definition of the genus some- 
what more precise than that given above, including certain peculiarities 
which seem likely to prove specific rather than generic. If they are re- 
tained the form described below and Siphonactinia Boeckii would be ex- 
cluded from the genus, to which they seem naturally referable. ! Rather 
than establish a new genus for their reception, I prefer to extend 
somewhat the limitations of the genus Peaenta. 

4. Peachia koreni, sp. nov. 
PI. XXI, Fig. 16. 

No. 954. Station, 2764. Lat" 36。42'S" long" 56。 23' W. Depth, Hi fathoms. One 

The single specimen of this species (PI. xxi, Fig. 16), which I dedicate 
to Prof. Koren, to whom, in collaboration with Prof. Danielssen, we owe 
the Fauna Litoralis Norvegise, is evidently closely related to P. (Siphon- 
actinia) BoecMi (Kor. et Dan.). I regret that I can not give as complete 
a description of it as I should like to do, owing to a disincliDation to 
mutilate the sole example obtained. 

Digitized by 


The base does not seem to have been adherent, but it is somewhat 
mutilated, so that it is not possible to be certain of this. No distinction, 
however, into capituliim, scapus, and pliysa is possible. The column 
is considerably wrinkled by contraction and shows no trace of tubercles 
or warts, and is not covered with foreign substances. Toward its 
lower part longitudinal grooves, marking the insertions of the mesen- 
teries, are to be seen, but they can not be traced upward toward the 
margin for any distance. The height of the column is and its 
diameter 0.8细. 

The margin is simple, and in the contracted specimen covers the bases 
of the teutacles. These are only eight in number and are short and 

The conchula, formed by the prolongation of the lips of the single 
siphonoglyphe, is as 】ong as the tentacles. On each side of the main 
portion of the conchula is a lobe rising only to about half the height 
of the former, and at the sulcular extremity of the mouth is a still 
smaller unpaired lobe. 

By cutting across the column until it was almost divided I was able 
to ascertain the arrangement and number of the mesenteries without 
appreciably mutilating tlie specimen. There is only one siphonoglyphe, 
which is long and deep^ with thick and firm walls, almost cartilaginous 
in their consistency. The mesenteries are twenty in number, arranged 
in ten pairs; two of these are directives, and in addition to these there 
are four other perfect pairs of about equal width, making altogether a 
principal cycle of six pairs of mesenteries. The remaining four pairs 
are imperfect and much narrower, and are situated in the sulco-lateral 
and lateral intermesenterial spaces. The longitudinal muscles are 

The arrangement of the mesenteries is the same as that found in 
Peachia hastata, but, as already stated, the general appearance of tbe 
animal, the form of its tentacles, and the possession of a well-devel- 
oped conchula bring it very close to P. (Siphonactinia) Bceckii. Whether 
the latter has also ten pairs of mesenteries remains to be seen. It has 
twelve tentacles, which would lead one to suppose that it was he'xam- 
erous, but the species here described shows, as does also Peachia has- 
tata with twelve tentacles, how little can be ascertained as to the num- 
ber of the mesenteries from the number of the tentacles. It is possible 
that the specimen of P. Koreni examined was young and had not de 
veloped its full quota of tentacles. I can not make any statements with 
regard to the presence or absence of reproductive elements, not having 
made microscopical preparations of the mesenteries. 

I think, however, that there can be no doubt as to the specific dis- 
tinctness of this species from tliat obtained on the Norwegian coast. 
The form of the conchula is entirely different, a fact in itself sufficient, 
in the present state of our knowledge of the anatomy of the conchula 
bearing Halcampidoe, to warrant the establishment of a distinct species. 
Proc. M, 93 —— 10 

Digitized by 


Family ANTHEAD^. 

The limitations of this family proposed by K. Hertwig ('82) seem sat- 
isfactory and will be adopted here. 

It seems doubtful whether such definitions as Andres ('83) proposes 
can be maintained for the genera Actinia and Anemonia, It may be, 
perhaps, better to unite all the forms of these genera which possess 
acrorhagi under the genus Actinia, leaving those destitute of such 
structures and without a distinct collar and fosse in the genus Anemo- 

Synonym : ComacHs flagelliferaf Hert. (non Dana). 
Nob. 957, 1739, Abrolhos Islands. Two specimens. 

The resemblance of these forms to that described by Hertwig as 
Comactis flagellifera is very great, and it seems almost certain that they 
are identical with it. They are somewhat smaller, measuring 0.25«™ in 
height, with a diameter at the base of 0.5^™, and at the margin of 0.7c™. 
The sphincter mnscle differs from the figure giveu by Hertwig ('82) 
only in being a little broader, and the radial muscles of the disk have 
the Cerianthan appearance which Hertwig mentions. The sipbono- 
glyphes are not well defined though easily made out in sections, and 
have as usual directive mesenteries counected with them. 

All the mesenteries appear to be perfect, though the youngest cycle 
are clearly marked out in sections below the middle of the stomato- 
daBum by being much narrower than the mesenteries of the other cycles, 
all of which are about the same width, so that a pair of broad mesen- 
teries alternates regularly with a narrow pair. I found indications of 
reproductive organs, but the ova were few in number, though fairly 
larfire, and appeared to occur in a few of the larger perfect mesenteries. 
Hertwig did not succeed iu finding reproductive elements in the speci- 
men he examined, and considered it therefore to be immature. Since, 
however, the specimens which I have studied possess ova and yet are 
smaller than Hertwig's specimen it seems probable that the latter is to 
be considered mature. 

One interesting histological peculiarity I have observed in this form 
may be mentioned. It is in connection with the structure of the upper 
part of the mesenterial filaments. The mesogloea has as a rule only a 
few scattered cells, but in the processes which support the median and 
lateral portions of the filament in its upper part the cells become ex- 
ceedingly numerous, closely packed together in the thickened mesogloea. 

Genus ACTINIA, Linn. 

'. Actinia infecunda, nom. nov. 

Plate XXI, Fig. 17. 


(PI. XXI, Fig. 17, mg), I have not met with such an arraugeinent in 
any other forms, and it forms a very striking peculiarity. 

As already stated there seems little room for doubt but that this 
species is the same as that described by Hertwig' as Comactis flagel- 
lifera. The external appearance is the same and the anatomical pe- 
culiarities are so similar that I do not believe a separation of them 
would be justifiable. Nevertheless, I have not followed Hertwig in 
his identification of the form. It was with some hesitation that he 
associated his form with Dana's Actinia flagelUfera, recognizing the 
great difference between his speciineu and the figure giveu by Dana 
('46). He regarded VerrilPs account ('66) of the alcoholic specimeDS as 
furnishing a reconciliation of the discrepaucies, iiotwithstauding the 
paucity of the facts which Verrill contributed. Johnson ('Gl) has, 
however, studied the sea anemones of the region where Dana's form 
was collected, nswnely, Madeira, and convinced himself that it was in 
reality identical with the common European Ane^nonia sulcata, which 
view is accepted by Andres* For this reason it seems advisable to 
separate Hertwig's Gomactis under a new name. 

(See Appendix P.) 

Plate XXI, Vi^s. 18, 19. 

No8. 694, 1362. Station, 2768. Lat., 42。 24' S. ; long" 61° 38' 30" W. Depth, 43 fath- 
oms. Numerous specimens. 

The numerous specimens (PI. xxi, Fig. 18) were for the most part only 
partially contracted, and measured in this condition 0.5 to 0.7**"^ in height 
and 0.5 to 1。™ in diameter at the base. They were seated upon sponges 
or occasionally upon Tubulariaii stems, the base of the Actiiiians in the 
latter cases surrounding]^ the stem. 

The column is somewhat wider at the base than higher up and has 
therefore a slightly conical shape. The ectoderm for the most part 
lias been macerated away, leaving the slightly translucent inesogloea 
exposed, and allowing the insertions of the mesenteries to be seen 
through the wall as fine 'longitudinal striae. The mesoj^loea is com- 
paratively thin and almost perfectly homogeneous, containing very few 
mesogloeal cells. No verruc je or acrorhagi are present. 

The tentacles are short and numerous, usually aopi oacliiDg one hun- 
dred, but vary ill g in uumbeiin the various specimens. One tentacle com- 
municates with each exo ― and each eiidocoel, and their uumber depends 
upon the number of mesenteries present in any one individual. In the 
majority of cases they are completely exposed, the sphincter muscle of 
the column being endodermal and diffuse and very weakly developed, 
as is the case with the general musculature throughout the body. The 

Genus A.NEMONIA, Risso. 

6. Anemonia variabilis, sp. no v. 


ectodermal muscles of the tentacles and disk form a simple layer, the 
mesogloea not being raised into supporting processes. 

The stomatodiBum is elongated, but without well-marked siphono- 
glyphes. Its ectodermal lining is thrown into very pronounced folds, 
supported by delicate though high longitudinal ridges of mesogloea 
(PL XXI, Fig. 19). 

The mesenteries are irregular in number. In sections of three speci- 
men for example, there were respectively 28, 33, and 30 pairs. As a 
rule a perfect and an imperfect pair alternate, but this amingement is 
not infrequently interrupted by the succession of two pairs of perfect 
ones, or of three or two pairs of imperfect ones. There are two pairs of 
directive mesenteries, and the number of mesenteries intervening be- 
tween tliem on each side is usually tlie same, though there are excep- 
tions to this rule. In the specimen of which a section is figured on PI. 
XXI, Fig. 19, it will be seen tbat only eleven pairs of mesenteries inter- 
vene between the two directives (7>) on one side, while there are as 
many as twenty-one on the other siae. This section represents the con- 
dition as seen towards the level of the lower extremity of the stomato- 
dseum. Higher up two pairs of mesenteries arc to be found which are 
not represented at tlie level figured, and these increase the number of 
mesenterial pairs of one side of the body to fourteen 一 i. e" thirteen 
pairs intervene between the two directives. Even in the uppermost 
sections, however, there is not equality in the number of the mesente- 
ries of either side. That the irregularity which is found in the succes- 
sion of perfect aud imperfect mesenteries is not an artificial production 
is shown by the relation of the perfect pairs on either side of the two 
(x) and three (y) imperfect pairs of the figure. It is there seeu that 
these nerfect pairs are attached to the stoinatodaeum opposite succes- 
sive inesogloeal ridges, and this relatioa of the ridges to the insertion 
of mesenteries, though not constant, is of sufficient frequence to war- 
rant the assumption that the groups of mesenteries y are truly im- 

The mesogloea of the mesenteries is considerably thicker a short dis- 
tance from their insertion into the column wall than elsewhere and is 
raised into only very low muscle processes. Consequently the muscle 
pennons are almost wanting, the longitudinal muscles forming little 
more than a simple layer over the surface of the niesoglcea. None of 
the specimens examined were mature; immature ova were observed, 
however, iu the endoderm of some of the perfect mesenteries and in that 
of the directives. 

The habits of this form suggested identity with that described by Ver- 
rill ('83) as Sagarita spang leola. Examination of specimens of the 
latter showed at ouce that the two forms were very different, S. spong- 
icolaj for example, possessing strong muscle pennons on tbe mesenteries 
attached by a slight pedicle in a manner recalling the conditions de- 
scribed by Hertwig ('82) for Leiotealia nymphwa. 

VOL. XVI, 1 
1893. J 



7. Anemonia (?) inequalis sp. nov. 

Plate XXXIV, Figs. 114 一 115. 

No. 742. PichiUngiie Bay, Lower (California. Littoral. Two specimens. 

The two specimens which represent this species are contracted, though 
the tentacles are not completely concealed. The base was adherent. 
In height the largest specimen measures 0.7*^*", with a diameter of 1.3。™. 
The column wall is thin and soft to the touch, and shows 72 longitud- 
inal lines which mark the insertion of the mesenteries. The ectoderm 
is completely macerated away. The mesoglcea is fairly thick and is, 
homogeneous in appearance, with numerous cells scattered tbrougli the 
matrix. A spliiiicter is preseut; it is endodermal and of the " diffuse" 
variety, formiufir, however, a not very compact mass aud being rather 
weak. (PL xxxiv, Fig. 114.) 

The tentacles are short, and apparently thirty-six in number, ar- 
ranged in a single cycle. Their ectodermal muscles are weak and are 
not embedded in the mesogl (お a. 

The stoaiatodaeuin is ridged longitudinally and possesses at least one 
shallow siphouoglyphe. In half the circumference of one specimen ex- 
amined eighteen pairs of mesenteries were present, from which it may 
be concluded that there are altogether thirty-six pairs, a number which 
corresponds with the number of longitudinal lines seen from the 
outside. Their arrangement is very peculiar. All are perfect above, 
but below they are evidently divided into three cycles, each consist- 
ing of twelve pairs. If the first cycle be considered to represent two 
primitive cycles, the apparent second cycle will really represent the 
third cycle, while the apparent third will be the fourth, in wliicli, how- 
ever, only half the proper number of pairs lias developed (PL xxxiv, 
Fig. 115.). The mesoglcea of the mesenteries resembles that of the 
column wall, being homogeneous and tolerably thick. The longitu- 
dinal muscles are not very strong and cannot be said to form a cir- 
cumscribed penuon. The parieto-basilars form folds upon the surface 
of the mesenteries, the edge ot the fold sometimes, however, uniting 
with the mesentery aud so producing one or more cavities enclosed 
within the mesogloea of the ingsentery near the insertion into the 
column wall. No ripe reproductive elements were present, but I suc- 
ceeded ill fiiuling a few very young mother cells, the macerated con- 
dition of the internal parts preventing, however, au accurate determi- 
nation of tlieir distribution. Some certainly occurred on one of the 
mesenteries of the second actual cycle and I thought I could distin- 
guish others on some mesenteries of the third aud fourth cycles, bat 
of this I can not be certain. 

1 assign this form provisionally to the genus Anemonia, It differs 
materially, however, from the typical forms of the group, as, for instance, 
in the short aud not numerous tentacles. The abnormal arrangement 


of the mesenteries is not, I believe, of sufficient importance to be 
generic and in the general structure there are undoubted affinities to the 
Aiitlieadie. As to the presence of acrorhagi nothing can be said, on 
account of tlie absence of the ectoderm, and the macerated condition of 
the internal parts proved a decided obstacle to a thorough study of 
the specimens. 

GemiH CONDYLACTIS, Duch. et Mich. 

The genus Condylactis was established in 18G6 by Duchassaing and 
Michelotti ('60) for the reception of the common West Indian form 0. 
passijlora. I have shown elsewhere ('89) tliat this form is in all re- 
spects an Aiithead, and that it agrees closely in general characteristics 
with the form described by Delia Chiaje as Actinia a/urantiaca, subse- 
quently assigned by Andres ('83) to the genus CereacUSj which is re- 
ferred to a special family. Tbe generic name proposed by Duchassaing 
and Michelotti has undoubted priority and must replace that proposed 
by Andres. I see no good reason for separating Condylaetis from the 
other Autheads, from which it is distinguished by the absence of acror- 
hagi and by the presence of a fosse between tbe margin and the bases of 
tbe tentacles, as well as by tbe usual presence of minute verrucae upon 
the column wall. 

8. Condylaetis cruentata (Dana). 
Plate XXI, Figs. 20—21. 

Synonyms : Actinia cruentata j Dana (1846) ; Cereua cruentatusy Milne-Edwards (1857) 
Bunodes cruentata, Gosse (1860). 

No. 736. Sandy Point, Straits of Magellan. Littoral. Four specimens. 

All the specimens (PJ. xxi, Fig. 20) are contracted, the tentacles being 
concealed ; in tliis condition the heiglit and diameter of the column are 
about tbe same .05 ゆ'. The preserved specimens show no coloration, 
but in sections brown granules of pigment are found in the endoderm 
of the disk and teiitucles. 

The base is adherent. The column wall is thrown into strong folds, 
and toward its upper part are rows of verrucae to whicli particles of 
sand are strongly adherent. The verrucae cease at tlie well-marked 
margin, between which and tlie bases of the external tentacles there is 
a well-marked fosse, which is made especially evident in contracted 
specimens by being drawn down by tbo strong longitudinal muscles of 
tlie mesenteries. Circular muscles are developed upon the column 
wall but are wanting at the margin ; internally to this, however, a few 
small muscle processes are found which represent the sphincter. It is 
very weak and can have ouly little eftect iu producing tbe coucealiiient 
of tbe tentacles; this is mainly brought about by the longitudinal 
muscles of the mesenteries. 

Tbe tentacles are not very numerous j their longitudinal muscles, like 
the radiating muscles of tbe disc, are not imbedded iu tbe mesogloea. 

Digitized by 



The stomatoddeum possesses well-developed siphouoglyphes with 
smooth walls, the rest of the stoinatodaeum being longitudinally ridged. 
There are only sixteen pairs of mesenteries, all of wliic;li are perfect, 
eight losing connection with the stomatodaeum, however, sooner than 
the others. The longitudinal muscles are strongly developed, forming 
a strong peunon (PI. xxi, Fig. 21), and the parieto-basilar (pbm) forms 
a strong fold upon the surface of the mesenteries. The reproductive 
organs are borne by the mesenteries of the first cycle, with the exception 
of the directives. acontia are present. 

There is necessarily some doubt as to the correctness of this identifi- 
cation. The external structure agrees well with Dana's species, as does 
also the habitat; as to the coloration nothing can be said. In referring 
it to the genus Condylactis, I bave separated it widely from the genera 
to which it has previously been (assigned. The nature of the sphincter 
and the arrangement of the mesenteries indicate a relationship to the 
Autheadae, and of existing genera of this family, by its possession of 
verrucae, and of a fosse, and by the absence of acroi hagi, it comes near- 
est to Condylactis. It differs from the described forms of this genus in 
its size and in the prominence of tlie verrucae, but it seems advisable 
for the present to include it in the genus. 

No. 731a. Station 2839. Lat., 33。 08' N.; long., S(P 15' W. Depth, 414 fathoms. 
Many specimens. 

In looking over the collection soon after it reached me I noticed that 
in the bottle which contained the species described below as Paractis 
vinosa, there were a large number of examples of a form which, while 
resembling the specimens of P. vinosa in general form and size, yet dif- 
fered decidedly in color. On submitting them to anatomical examina- 
tion I found that very decided structural differences existed, and that 
1 had to do not only with a distinct species, but even with a member 
of a distinct family. After much uncertainty as to the family to which 
it should be assigned, I determined to insert it in this report as an ap- 
pendix to the Autheadse. My reasons for so doing will better be under- 
stood after a description of the specimens has been given. 

They are all more or less contracted, some having the tentacles com- 
pletely contracted, while in others they remain more or less exposed 
(PI. XXI, Fig. 22). The color of the column and tentacles is pale pink 
or flesh color. In height the less contracted specimens measure about 
lcni, their diameter ranging from 1.3 to 1.5彻. 

The base is adherent, and in many specimens is more or less covered 
by a dark brown, somewhat granular cuticle. Its diameter is as a 
rule somewhat greater than that of the column ; in the specimen from 
which the measurements given above were taken its diameter was about 

9. Myonanthus ambiguus, gen. et sp. nov. 

Plate XXI, Fig. 22; Plate xxii, Fig. 23. 


2c™. . Its mesogloea is rather thin, allowing the straw-yellow color of 
the reproductive organs to shimmer through. 

The column wall is smooth for the most part, except for the slight 
folds caused by contraction. In the more intensely contracted speci- 
mens in the upper part twenty four longitudinal folds were more or less 
distinct, terminating abruptly at the margin ; twelve of the folds are 
smaller tliau, and alternate with, the other twelve. The ectoderm has 
been to a large extent macerated away from the column wall, but where 
present it has the same color as tho inesoglcea. No trace of verrucas or 
tubercles could be discovered. The mesogloea is much thicker than that 
of the base and has a fibrous structure. It is not, however, stiff and 
parclimeut-like to the touch, but ou the other hand rather soft and 
tough. Just at the margin, where the longitudinal folds of the con- 
tracted specimeus terminate, is a well-deveioped endodermal sphincter 
(PI. XXII, Fig. 23). It can hardly be classed either as "circumscribed" 
or "diffuse," since, though well defined, it is not connected to the 
column wall by a distinct pedicle. It is ratber intermediate between 
these two varieties of sphincter, and resembles closely that form of 
muscle wiiicli I have elsewhere ('80 a) described for a species of Phyl- 
laetis. I would suggest the application of tlio term "aggregated" for 
this variety of muscle. Its appearance in cross section may be under- 
stood by a reference to PI. xxii, Fig. 23. It is to be observed that anas- 
tomes between the muscle processes are not uufrequeut, so that bun- 
dles of muscle fibers become enclosed within the inesoglcea. Imme- 
diately adjacent to the sphincter the ordinary circular muscles of the 
column are hardly developed, but lower down tbey become stronger, 
without, however, forming a second sphincter. 

The sphincter seems to occur just at the margin, and apparently a 
slight fosse exists between this and the outermost tentacles. No acro- 
rhagi could be detected. The tentacles are uuinerous and arranged in 
several cycles, but I could not ascertain their actual number. Their 
color is the same as that of the column and they are of fair length aud 
rather slender. Their longitudinal musculature is not imbedded in the 
mesogloea, resembling in this respect the radial muscles of the disc. 

The stomatodiBum possesses two siphon ogly plies whose lower ends 
are prolonged apparently some distance beyond the lower openiDg of 
the stomatodiBum. In consequence of this tlie directive mesenteries 
are attached to the stomatodieum throughout a greater extent of their 
length than are the other perfect mesenteries. Tliere are, altogether, four 
cycles of mesenteries, of which only the six pairs forming the primary 
cycle are perfect. The mesenteries of the fourth cycle are small, and 
hardly project beyond the eudoderin. All the mesenteries except those 
of the fourth -cycle and the directives are goiiophoric. Tbe reproduc- 
tive organs are very evident in dissected specimens, owing to their 
bright orange color due to tbe presence of numerous oil globules in tbe 
ova and sperm mother cells. The mesenterial filaments are not deeply 
colored, as in P. vinosa. This forms a simple point of distiaction be- 


tweeu the two associated species. The longitudinal muscles are only 
moderately developed, and do not form a well-marked pennon. The 
parieto-basilar seems to be weak. No acontia occur. 

From the above description it may be seen that this form is not read- 
ily referable to any of the recognized families. Ou the whole, however, 
it seems to approach more nearly the Antbeadae than any of the others. 
The smooth column wail and the distribution of the reprod ひ ctiv^e ele- 
ments are points oi similarity, but on the other haud the small number 
of perfect mesenteries aud the strong sphincter are decided differences. 
The sphincter, however, is practically an eiidodermal cue, and the 
definition given by Hertwig ('82) for the Aiitlieadje does not exclude 
the existence of a recognizable sphincter. In fact, in Actinia infecunda, 
which he recognizes as au Anthead, a spbiucter is present of such a 
form that an excessive amount of differentiation would not be called into 
play to transform it into such a muscle as we find in Myonanthus. 

I think, accordingly, that it is advisable to refer this form to the family 
Antheadae, -regarding it as a somewhat aberrant form, wliich has the 
power of completely retracting the tentacles, owing to the possession of 
a well-defined sphincter, a character which has suggested the generic 
name I have applied to it (fiuwv = a knot of muscles). 


Actininae with usually stout ^lonretractile tentacles, strongly con- 
stricted immediately above their insertion into tlie disk, and hence 
readily deciduous. Sphincter muscle endodermal, diffuse, or in some 
forms approaching the circumscribed type; the tentacles and disk fully 
exposed in the contracted condition. With more than six pairs of per- 
fect mesenteries. 

Bolocera tuediw discovered, iii 1832, by Jolinston, and later referred 
by him ('47) to the genus Anthea may be taken as a typical example 
of this family. Gosse ('60) established for it the geims Bolocera, and 
separating it from the Aiitheadae, with which Johnston and Milne-Ed- 
wards ('57) associated it, placed it among the Bunodidae, in which clas- 
sification he has been followed by Andres ('83). A study of the form 
occurring in the deep water off the eastern coast of tlie United States, 
and which lias been identified by Prof. Verrill witli B. tuedice, as well as 
of other species of Bolocera from the Albatross collection, has deinou- 
stratedthat, so far as their anatomical peculiarities are concerned, these 
forms are very different from the Buuodidiv, but stand in relatively 
close affiliation to the Antheadae. The nature of the tentacles, Low- 
ever, and other structural characters, seems to be of sufficient import- 
ance to warrant the establishment of a distinct family for them.* 

*It seems probable that Danielssen's ('90) Sideractis is a BoJocera, though the ex- 
istence of an endodermal sphincter would preclude such an identification, li is to 
be noticed, however, that Danielssen's figure (PI. vii, tig. 10) hardly bears out his 
assertion on this point. 

Digitized by 


Genus BOLOCERA, Gosse. 

With the characters of the family. 
10. Bolocera occidua, sp. nov. 
Plate XXII, Figs. 24-27. 
No. 706. Station 2783. Lat. 51 し、 02' 30" S. ; long. 74^ 08' 30" W. Depth, 122 fath- 
oms. Two specimens. 
No. 701. Station 2779. Lat. 53。 06' S. ; long. 70。 40' 30" W. Depth, 77i fathoms. 
Three specimeus. 

No. 697. Station 2771. Lat. 51。 34' S.; long. 63° W. Depth, 50^ fathoms. Two 
specimens (young). 

The base is evidently adherent, and is slightly smaller than the col 
umn. It is marked with fine radiating ridges, which are continued 
over the linibus upon the column. 

This is nearly cylindrical, expanding slightly above, and in the con 
tracted condition can not be said to possess verrucie or warts, though 
the entire surface is marked out into small quadrangular areas by the 
crossing of vertical and circular furrows, processes of mesogloea sup- 
porting the ectoderm of the elevated areas (PI. xxii, Fig. 24). In tlic 
largest specimens the height and diameter of the column are about 
3cni. Near the margin, in most of the specimeus, complicated structures 
could be seen which, on examination, proved to be mesenterial filaments 
protruding from openings formed by the falling off of the tentacles. 

The margin is tentaculate. The tei^acles are large and stout, cover- 
ing the greater portion of the disk. They are arranged iu about four 
cycles, of which the two inner cycles each possess twelve tentacles, the 
third cycle twenty-four, and the fourth forty-eiglit. The tentacles re- 
tain their cylindrical shape in the preserved specimens and are plainly 
furrowed (PI. xxii, Fig. 26). At their insertion into the disc tbey sud- 
denly diminish in diameter, so that they are attached by a short and 
narrow pedicle; tbey are thickest immediately above the pedicle, where 
the inner tentacles in the largest specimen measured 0.9c™ in diameter, 
and from that taper gradually towards the extremity, which is some- 
what obtuse. The length of the tentacles of the innermost cycle in the 
largest specimen was 5咖. 

In COD sequence of their manner of insertion into the disc the tentacles 
are readily broken off, leaving a circular opening upon the disc which 
indicates their former position. The opening, however, is diminished 
by a circular fold of niesogloea, covered by endoderin, which encroaches 
upon it (PI. XXII, Fig. 27, tsp); the free edge of the fold is thrown into 
numerous muscle processes, and it seems probable that by the approxi- 
mation of the edges of the fold the opening may be completely closed.* 

* Since this was written Carlgren ('91) has described, in a paper on B. longicornis, 
a similar sphincter fold. He points out, correctly, that the sphincter is thrown off 
with the tentacle, and it therefore does not serve to close the opening left on the 
surface of the disk. My description was drawn up from sections through tentacles 
still adherent, and the conclusion was somewhat hastily reached that the use of the 
sphincter fold was to occlude the opening. 

Digitized by 



Tlie circular muscles of the tentacles aud disk are ectodermal and com- 
paratively weak. 

The lips are prominent, aud are marked by delicate aud numerous 
striae, which are continued down the stomatodaeuin and apparently 
correspond approximately to the mesenteries. Two siphonoglyphes 
are present and are deep, the directive mesenteries being comparatively 

The sphincter muscle (PI. xxii, Fig. 24, sph) is enclodermal and diffuse, 
the endodermal muscle processes of the column being more numerous 
and somewhat higher just below the margin than elsewhere. 

The mesenteries are arranged in three cycles. The first cycle con- 
sists of twelve perfect mesenteries, including two directives, the sec- 
ond cycle likewise of twelve mesenteries, which are imperfect, bow- 
ever, and the third cycle of twenty-four mesenteries, which are quite 
narrow and imperfect. All the mesenteries of the first aud second 
cycles, with the exception of the directives, are gonophoric. The Ion 
gitudinal ra ひ scles of the mesenteries are fairly well developed, the 
supporting process covering the entire nou-gouophoric regiou of the 
mesentery and being of almost uniform height throughout. (PI. xxii, 
Fig. 25.) There is no special development of the parieto-basilar imiscle. 

In its general appearance B. occidua resembles very closely B. tue- 
dice. I have been able, however, to examioe some preserved spe (! i- 
mens of the latter obtained from the deeper water off our eastern coast, 
and can state that there are marked dift'erences in the anatomy of the 
two species. For instance, B, tuediw has tbe tentacles arranged in 
only three cycles, aud the parieto-basilar muscles upon the mesenteries 
show a condition similar to what occurs in B. parmom, to be described 

It is possible that the form here described may be identical with 
Studer's ('78) B, Jcerguelensis, which is described as having the ten- 
tacles arranged in several cycles. We possess, however, no account of 
tbe auatomical peculiarities of this form; aud since the general shape 
of the body differs decidedly from that of B. occidua, and there are 
said to be seven cycles of tentacles in large specimens, I have consid- 
ered it advisable to separate tbe two forms. I believe that in a case of 
doubt it is preferable to consider the newer form a distinct species ; 
the union of forms improperly separated is a mucli simpler matter tliau 
the separation of forms em^neously identified. 

A third forui, with which B, occidua might possibly be ideutified, is 
B. multicornisj of Yerrill ('79). Andres ('83) places this form among 
the doubtful Buuodidiie, not being ab】e to determine from YerrilPs de- 
scription whether it is truly a Bolocera or uot. I have been able to 
examine a specimen of it, however, and can confirm Verrill's assign- 
ment of it to that genus. The greater number oi its tentacles aud 
their much smaller dimensions show that it is distinct from B, occidua. 


11. Bolocera pannosa, ap. noy. 

Plate XXII, Figs. 28 and 29. PI. xxiii, Fig. 30. 

No. 729. Station 2839. Lat. 33。 08' N. ; long. 118° 40' W. Depth, 414 fathoms. 
Eight specimens. 

This form, in its preserved condition, presents at the first glance only 
a remote similarity to other species of Bolocera. One misses the ro- 
bust appearance and the large, stout tentacles, and finds instead a 
ragged mass. Closer observation reveals, however, many points of sim- 
ilarity to B. tuediWj and it is necessary to consider botli as belonging 
to the same family, and probably also to the same genus. 

The base is oval and attached. In average specimens it measured 
7復 in length and 2.5續 in breadth. It is thin, especially toward the 
center, allowing tlie mesenteries and the dark, wine-colored pigment 
of the mesenterial filaments to be indistinctly perceived. Toward the 
periphery radiating and concentric grooves are readily made out, mark- 
ing off the surface into small quadrangular areas. 

The column is low; in noue of the specimens does it exceed 0.7。™ in 
height, and it is folded back upon itself, so that the laargiu and liinbus 
are nearly in coutact. Immediately below the region where the bending 
back occurs is a relatively strong, circumscribed endodermal sphincter, 
which is, no doubt, the cause of tlie reversion of the inargin. This 
sphincter (PI. xxiii, Fig. 30) consists of a mainmesogloeal process project- 
ing out almost at a right angle to the column wall and giving rise to 
numerous secondary processes mainly on its marginal side, other proc- 
esses arising below it directly from the column wall and grading off 
into the ordinary circular muscle processes. Tliis sphincter, it will be 
noticed, is situated low down on the column wall, some distance away 
from the margin. Muscle processes supporting circular muscles occur 
above it, but they are not specially aggregated to form a sphincter. 
The sphincter which is present is to be regarded as a lower sphincter, 
the marginal sphincter not being developed. 

The surface of the column is divided into small quadrangular areas 
by longitudinal and circular lines corresponding to the radiating and 
concentric grooves of the base. No warts or verrucae, however, seem to 
be present, nor are there any very decided inesogloeal processes sup- 
porting the quadrangular areas as in B, oecidua. 

The entire animal is of a pale rosy tint, or in some cases salmon- 
colored, the mesenterial filaments being of a deep wine purple. Prob- 
ably in life the colors were more pronounced, resembling the coloration 
which seems usual to the Boloceridae. 

The margin is tentaculate. The tentacles are numerous and strongly 
entacmaeous, arranged apparently in about seven cycles, 6, C, 12, 24, 
48, 96, 192. The inner tentacles measure about 3.7*^™ iu length, and 
apparently are not capable of being contracted to any very great ex- 
tent. In their general structure the tentacles resemble those of B, tue- 


dice, being constricted just at their insertion into the disc, being widest 
just distal to the constriction ; consequently they readily fall off, leaving 
a round opening in the disc. These openings are, however, partially 
closed by a muscular fold of mesogloea arising from their margins, and 
similar to what has been already described for B, occtdua.* ^^otwith- 
standing their close similarity in the structure, the tentacles have a 
very different appearance from those of the species just mentioned. In- 
stead of being plump, turgid, and robust, they are flaccid, flattened, 
and ratlier slender, and give to the preserved specimen a very ragged 
and torn appearance. It is on this account that I have bestowed upon 
the species the name pannosa. 

The musculature of the tentacles is weak, the ectodermal muscles 
not being imbedded in the mesogloea (PI. xxii, Fig. 28), but supported 
by hardly noticeable mesogloeal elevations. The longitudinal ridges of 
mesogloea which give to the tentacles of the BoloceridsB their fluted 
appearance are readily to be seen in the tentacles of the inner cycles, 
but they are not so well developed an in other species of Bolocera. 

The disc is almost entirely covered by the tentacles, only a relatively 
small area around the mouth being naked. Its ectodermal muscles 
are weak, though the endodermal circular system is fairly well devel- 
oped 5 less so, however, toward the margin. 

The stomatodseura is prominent and possesses two siphonoglyphes. 
The mesenteries are numerous, there being probably about ninety-six 
pairs,, of which twenty-four are perfect and non-gonoplioric (two of them 
being directives), twenty-four well developed, though not perfect, and 
forty-eight relatively small. All the imperfect mesenteries are gono- 
phoric. The musculature of the mesenteries is not particularly strong, 
but presents a very peculiar arrangement. If a transverse section of 
a mesentery of the first cycle be examined (PI. xxi, fig. 29) it will be 
seen that at its attachment to the column wall it is comparatively thin; 
it soou, however, becomes thicker, and numerous cavities, containing 
apparently the degenerated remains of cells, are seen iu the mesogloea. 
The exoccelic face of this portion of the mesentery bears muscle proc- 
esses which are cut transversely (pbm)j and therefore give support 
to longitudinal muscles, or rather to the oblique muscles forming the 
parieto-basilar muscle. The inner edge of this iiiuscle is to a slight 
extent free from the nieseutery, and it seems as if the cavities had been 
produced by the fusion at intervals of the mesoglceii of this free edge 
with that of the mesentery diilinj? the growth of the animal. Beyond 
tlie region of the parieto-basilar muscle the mesogUpa becomes tbinner, 
Hud its exoccelic surface is covered by a simple layer of muscle cells 
whose fibres internal to the parieto basilar region ran longitudinally, 
then became transverse, and finally near the insertion of the mesentery 
into the stomatodteiim become again longitudinal, being now supported 

See note p. 154, 


on short processes of mesoglopa. On the endocoelic face of the mesen- 
tery near its insertion into the column wall are muscle processes bear- 
ing longitudinal muscles (Im), but the greater portion of the surface is 
covered by a well-marked layer of transverse muscles (《m), amongst 
wbicli, however, some longitudinal fibres may be detected. This trans- 
verse layer covers about two-thirds of the surface, but the third adja- 
cent to the stomatodaBum is occupied by the moderately developed 
longitudinal muscle-pennon (Im), The arraDgeraent appears at first 
sight to be tlie normal relations reversed, so far as the faces of the 
mesentery are concerned, and to a certain extent this is the case. 
The greater portion of what normally would be exocoelic transverse 
musculature has become loiigitudinal, while the endocoelic longitudinal 
musculature has to a large extent become transverse. The longitudi- 
nal inuscle-pennon, and the parieto-basilar muscle still, however, retain 
their normal relations. 

A histological point was well shown in the preparations of this form, on 
account of the specimens having undergone a certain amount of macera- 
tion in the preserving alcohol. Delicate mesogloeal filaments can 
readily be seen to exteud from the iimscle processes out between the 
cells, both of the ectoderm and the endoderm. I have called attention 
to this fact ID the case of Cerianthus americanus (,90), and have since 
observed it in numerous forms, so that it is probably a normal arrange- 

12. Bolocera brevicornis, sp. no v. (ト' ee Appendix, p. 209.) 
PI. XXIII, Figs. 31-33. 

No. 730. Station 2839. Lat. 33^ 08' N., long. 118^ 40' W., 4U fathoms. Two speci- 

This interesting form was dredged in the same locality as B, pannosa. 
It is represented in the collection by two specimens, one of which is 
apparently full grown, while the other is evidently young. The base 
is circular in outline and adherent. It measures in the large specimen 

The column wall is bent downwards, so that the margin is almost level 
with the base, and the whole expanse of the disk is exposed. The column 
is marked by numerous longitudinal lines, extending from the lim- 
bus to the margiD, where tliey terminate in a well-marked circular fold. 
Apparently the unper portion of the column is furnished with verrucae, 
but owing to the somewhat imperfect preservation of the column ecto- 
derm it is impossible to be certain on this point. The mesoglcea of the 
column is moderately thick, and on its inner surface is richly folded, 
so that the circular musculature is relatively strong. In tbe region of 
the circular fold, which forms tbe margin, the muscle processes are 
longer and more closely aggregated than elsewhere, forming a well- 
marked endodermal sphincter of the diffuse type (PI. xxiii, Fig. 31). 
Below the sphincter the wall Ls thinner than elsewhere, and lias the ap- 

Digitized by 


1893. J 



pearance of being pouched, the pouches perhaps corresponding to ver- 
rncae. Below this thin region the muscle processes are somewhat 
longer than further down, suggesting a second sphincter. 

The disc is very broad, measuring 6。™ in diameter. Its whole sur- 
face, with the exception of a small area immediately surrounding the 
mouth, is covered with tentacles, or with openings which correspond 
to them. The tentacles must have been" exceedingly numerous when 
all were present, having been arranged in as many as fourteen or fif- 
teen cycles. They are short, very short, when compared with those of 
B. tuediWy those of the inner cycles, a few of whicli persist in the 
large specimen, measuring only l.G*"» iii length. In other respects, 
however, they have all the characteristics ot the Bolocerid tentacles. 
They are attached to the disc by a narrow neck, the mesogloea of which 
is very thin. They are readily deciduous and they are fluted. In 
character they resemble the tentacles of B. pannosa rather than B. 
tuedicBj being somewhat flaccid. Above the neck of the tentacle there 
is a sphincter-bearing fold of mesogloea, projecting into the cavity of 
the tentacle, as in other Bolocerids. 

The mouth is slightly prominent and two well developed siphono- 
glyphes are present. It is difficult to estimate the number of mesen- 
teries present. I judge that there are about forty-eight pairs of perfect 
mesenteries. Between each pair of perfect mesenteries there are three 
well defined series of meseutei iea of gradually diminishing size and be- 
longing to three different cycles, so that if the estimate of forty-eight 
is correct for the first cycle, there will be in all three hundred 'and 
eight-four pairs of mesenteries, arranged in four regular cycles. This 
number does not, however, at all compare with the number of tentacles, 
and if the column wall be (ilosely examined a number of minute ridges 
may be seen between the pairs of mesenteries, hardly, if at all, rising 
above the level of the endoderm, and not apparently arranged in regular 
pairs or separable into definite cycles. These seem to be somewhat 
irregularly formed abortive (or incipient f) mesenteries, an attempt be- 
ing apparently made to preserve the relation of mesenteries to tenta- 
cles which is usually found. 

The specimens examined show no trace of reproductive organs, but 
from the general appearance of tbe mesenteries it is presumable that 
the ova or spermatozoa are borne by the imperfect mesenteries of the 
second, third, and fourth cycles. 

The musculature of the mesenteries is weak and presents no such pe- 
culiar appearance as has been described for B, pannosa. Tbe muscles 
on the endocoelic face, however, appear to be transverse in the region 
near the column wall, but form a low and diftuse longitudinal muscle 
pennon covering the inner three-quarters of tbe iimscle-bearing region 
of the mesentery (PL xxiii, Fig. 33). The parieto-basiJar 】imsde is 
present (PI. xxiii, Fig. 32), as shown by the direction of its fibres, 
but it produces no such cavities in tbe mesoglcea of the region of nies- 


entery occupied by it as it does in B. pannosa. The musculature of the 
lest of the excoelic surface is for the most part oblique, becoming for 
a short distance transverse, aud finally, as in B. pannosa^ becoming lon- 
gitudinal. The general arrangement of the musculature therefore 
agrees closely with that of B. pannma^ the main difference being the 
absence of cavities in the mesogloea of the parieto basilar region. 

This form is one of considerable interest. When I first saw it in glanc- 
ing over the collection, I believed I had before me a specimen of Hert- 
wig's Liponema mtiltiporum ('88). The presence of the tentacles, 
however, induced me to believe that I was wrong in this supposition, 
but the general similarity in appearance suggested the idea that pos- 
sibly Hertwig's specimens were identical with this, but had lost all their 
tentacles. When I had finished my study of the anatomy of B. hrevi- 
cornis, I perceived that this idea was not quite correct, but that though 
the two forms can not be considered identical specifically, yet they are 
so closely related as to warrant the conclusion that they belonged to 
the same genus, and that Liponema multiporum is a Bolocera which lias 
lost all its tenta<3le8. 

To anyone who has followed my description carefully and has 
compared it with that of Liponema^ I think the similarity between the 
two forms will be apparent. There is the same general appearance, the 
same folding back of the voluminous disk, the same "stomidia" almost 
covering the disk (though in the Albatross form these are normally sur- 
mounted by tentacles), the same circular fold at the margin, the same 
longitudinal lines on the column, a similar double endodermal sphincter, 
the two muscles being separated by pouchings out of the colurau wall, 
the same discrepancy between the number of mesenteries and tentacles 
(or stomidia), and a close similarity in the arrangement of the perfect 
and imperfect mesenteries. 

These similarities are, I think, sufficient to mark the two forms as 
belonging to the same genus. The different shape of the marginal 
sphincters and the slight difference in the arrangement of the mes- 
enteries leads to their assignment to distinct species. 

It is worthy of note, too, that Ilertwig describes a sphincter fold 
closing the openings on the disk, the "stomidia." This reminds me 
strongly of the muscular fold in the tentacles described in the preceding 
vSpecievS of Bolocera, Taking all the facts into consideration, I believe 
that Hert wig's Liponema muUiporum should bencefortli be kuown as 
Bolocera multipora. 

Actiniiie usually with numerous perfect mesenteries; circular muscle 
strong, imbedded in the mesogloea: acontia wanting. 

The family Paractidae was established by R. Hertwig ('82) on ana- 
tomical grounds, the forms belonging to it having been previously 
associated for the most part with the Antheadae. lu the above defini- 

Family PABACT【D^, R. Hert. 



tioD I have modified somewhat that given by Hertwig, thereby extend- 
ing the limits of the family so as to include certain forms with short, 
stout, non-retractile tentacles. I consider the presence of a strong 
mesogloeal sphincter and the absence of acontia the two most marked 
characteristics of the family, the number of mesenteries being of less 
importance, for although the majority of forms to be assigned to the 
family possess numerous perfect mesenteries, there are nevertheless 
some in which only the mesenteries of the first cycle are perfect. These 
are, however, so closely related to those with numerous perfect mesen- 
teries that it seems to me injudicious to separate them. 

Andres ('83) independently established a family Paractidae, which 
probably is iden;tical with that of Hertwig. The definition was, how- 
ever, founded altogether on external characters, which are undoubtedly 
of less value in Actiniau taxonomy than are anatomical features. 

Genus PARACTIS, M.-Edw. 

Paractidae with smooth body-surface, without papillae or marginal 
spherules; tentacles, slender, not exceptionally numerous, nearly equal 
in length and strength; margin not lobed. Sphincter widening some- 
what abruptly in its upper part, and occupying near the margin nearly 
the entire thickness of the mesogloea. This is the definition which Hert- 
wig ('82) gives of the genus, with the exception that he includes in 
the definition the presence of " numerous longitudinal furrows of the 
wall,'' which it appears to me limits the genus too narrowly, and by what 
is probably a more or less trivial character. He himself points out the 
possible alliance of his P. excavata to the Actinia peruviana of Lesson, 
in which the longitudinal fdrrows, are wanting, except near the base, 
the column wall being described as smooth. 

In the Albatross collection there are two forms which must be as- 
signed to the genus as here limited, although they differ greatly in 
certain respects. In one, the column wall, though not particularly 
thick, is leathery, while in the other it is of a much softer consistency; 
and again in one the radial muscles of the disc and longitudinal muscles 
of the tentacles are imbedded in the mesogloea, while in the other they 
are ectodermal. Whether this latter feature is one sufficient for ge- 
neric distinction can only be determined by the examination of a large 
number of Paractidse. I propose to place both the forms provisionally 
in the genus Paractis^ leaving it for future workers to decide as to the 
advisability of their separation. There is one feature in which they 
both agree, and that is in the shape of the sphincter muscle, which from 
being very narrow below gradually widens as it nears the margin, and 
has consequently a somewhat club-shaped form. Apparently P, exca- 
vata has a similar sphincter, though Hertwig has given no figure from 
which its form may be accurately determined, 
Proc K M, 93 —— 11 

Digitized by 


13 Faractis lineolata (Dana) M.-Edw. 

Plate XXIII, Figs. 34-36. 

No. 719. Station 2804. Lat. 8^ 16' 30" N.; loug. 79^ 37' 45" W. Depth, 47 fathoms. 
Eight specimens. 

The species to which I refer the form about to be described was first 
mentioned by Dana ('46) as Actinia lineolata, and was subsequently 
referred by Milne-Edwards ('57) to his geiras Faractis. Yerrill (,68), 
however, removed it from that genus and placed it in the genus Sagartia, 
and Andres ('84), assuming it to be a Sagartid, assigned it to Nemac- 
tis. In its general appearance the "Albatross ,, specimens seem to agree 
with Dana's description, and the absence of acontia show that they are 
to be replaced in the genus Par act", as it is here understood. 

The individuals are small (PI. xxiii, Fig. 34), and, for the most part, 
contracted to a hemispherical shape, the tentacles being entirely con- 
cealed, as a rule, though in some specimens they are not perfectly 
infolded. The base, which is adherent, measures iu the contracted 
specimens 0.5"", and the height of the contracted column is about 0.6c™, 

The column is pale in color and is marked with irregular chocolate- 
brown spots arranged distinctly in rows, and giving the effect of longi- 
tudinal bands of brown on a pale ground. There is some variation in 
the width of the bands, but I could not make out a regular alternation of 
three narrower bands with a wider one, sucU as Dana describes. The 
column wall is perfectly smooth ; Its mesogloea below is rather than, 
but near the margin it thickens rather suddenly. In this thickened 
region the sphincter muscle (PI. xxiii, Fig. 36) is imbedded. It occu- 
pies in its upper part nearly the entire thickness of the mesogloea, being 
separated from the endoderm on the one side, and tlie ectoderm on the 
other, by only a small band of mesogloea. In its lower part it tapers 
off, and lies nearer the endodermal than the ectodermal surface. The 
mesogloea throughout the column wall has a fibrous appearance, and 
the slightly oval muscle cavities appear to be separated by fine fibrous 
partitions in transverse sections. 

The tentacles are short and obtuse; iu one specimen in which tliey 
could be seen they were numerous, probably numbering ninety-six, 
while in another there seemed to be only forty-eight. In this respect 
the form here described differs from Dana's A. lineolata, which is de- 
scribed as having only twenty-four tentacles, arranged in two cycles. 
The loDgitudinal muscles of the tentacles, and the corresponding radial 
ones of the disc, are rather weakly developed and are entirely ecto- 
dermal in position. The tentacles seem to cover a large portion of the 
disc, though, owing to the contracted condition of the specimens, this 
could not be accurately ascertained. 

The mesenteries are few in number, and are arranged in four cycles. 
The six pairs of the first cycle are alone perfect, those of the second 
and thira cycles are goaophoric, while those of the fourth cycle are 



quite small and are destitute of mesenterial filaments. This was the 
arrangement in a specimen which had about forty-eight tentacles. It 
will be seen from this that we have an arrangement of the mesenteries 
which Hertwig considers typical for the SagartidsB, but a careful 
search, both in dissected specimens and in sections, for acontia failed 
to reveal their presence. The longitudinal muscles of the mesenteries 
form a distinct, though somewhat narrow, pennon (PL xxm, Fig. 35), but 
the parieto-basilar appears to be very weak. 

14. Faractis vinosa, sp. iiov. 

Plate XXIII, FiKH. 37-40; Plate xxiv, Fig. 41. 

No. 731. Station 2839. Lat. 33。 08' N. ; long. 118^ 40' W. Depth, 414 fathoms. 
Many specimens. , 

The majority of the specimens were contracted, iiiauy, however, show- 
ing the teutacles protruding (PI. xxiii, Fig. 37), while iu others they were 
not at all infolded. In the latter the height of the colunm was 1.4 to 
1.6c'" aud its diameter 1.2 to 1.(5^"'. 

The base is adherent and thin, allowiug: the iusertiou of the mesen- 
teries and the dark color of the mesenterial filameuts to be seen through 
it. In some specimens it is covered by a somewhat granular membrane, 
which is very friable and easily removed in fragments, and seems to be 
equivalent to the finn basal membraue oceuiTiiig, for instance, in Adam- 
sia palliata. 

The column is of a leathery cousisteiice, quite thin near the base, 
where it is marked with vertical furrows corresi>oudin g to the meseuteries, 
and fading out rapidly above. Iu color the column wall is white, owing 
to the absence of ectoderm, the few fragments of this which persist 
being of a pale brown color. The iiiesoglcea lias a finely granular ap- 
pearance in sections aud is thickest near the margin. The sphincter 
muscle (PI. XXIV, Fig. 41) occupies the greater part of this thickened re- 
gion and is strong. Below it tapers ott' slowly, extending a long dis- 
tance down the column wall, lying immediately below the endoderm 
and passing gradually into the ordinary eiulodermal circular muscles 
which are well developed and borne, on strong processes. (PI. xxiii, 
Fig. 40.) 

The margin is smooth, although in some more contracted specimens 
it may be thrown into a few folds. The tentacles are arranged in about 
four cycles, aud their number appears to be sixty-four They are white 
and translucent, but probably tbis is due to the ectoderm having been 
macerated away from their exposed surfaces, since in some of tlie 
stroDgly contracted specimens the ectoderm of the tentacles contains 
granules of reddish pigment. Tbe disc is of a deep wine color, as is 
also the stomatodiieum, the pigment granules being so abundant iu tbe 
ectodermal cell's as to completely obscure tbeir structure. The ecto- 
dermal muscles of tlie tentacles and disk are imbedded in the mesogloea, 

Digitized by 


occurring in the tentacles at about the middle of that layer. (PL xxm, 
Fig. 38.) 

The stomatodajuin is thrown into strong folds, borne on rather stout 
longitudinal elevations of the mesogloea. The siphonoglyphes are 
deep with smooth walls, and the ectodermal cells lining them have the 
pigment confiued to their outer ends ami not scattered through their 
entire thickness as happens elsewhere and on the stomatoddBum. 

The mesenteries are thirty-two in number, sixteen being perfect and 
sixteen imperfect. The longitudinal muscles are fairly well developed 
(PI. XXIII, Fig. 39), covering the greater portion of the surface of the 
perfect mesenteries; the parieto-basilar is not, however, particularly 
strong. Only the imperfect mesenteries are gonophoric, and the repro- 
ductive organs are very conspicuous on account of their bright orange 
color due to the presence of large oil globules in the ova and sperma- 
tozoa mother cells. The mesenterial filaments are, like the disc, of a 
deep wine color, the general eiidoderm being colorless. 

In its coloration, so far as this can bo determined, this form comes 
close to Paractis rubutt obtained by the Wilkes Exploring Expedition 
at Valparaiso. The very different habitat of the Albatross form, which 
is an inhabitant of deep water and the uncertainty of an indentification 
of ail alcoholic specimen with a form described as seen living and with- 
out any characteristic anatomical features, has induced me to consider 
for the present the Albatross form as distiuct. 

Paractidae with a large number of short tentacles covering the greater 
portion of the disc ^ margin of the disc lobed as iu Metridium. Sphincter 
strong, prolonged a long distance down the wall. 

Hertwig ('82) established this genus for a form previously referred 
to the genus Metridium, and whicli bears strong resemblance to the 
forms properly belonging to that group, at least in so far as the margin 
and the tentacles are concerned. Oii the other band, Hertwig has 
shown that in this case the external similarity is accompanied by such 
differences in internal organization that the establiaLraent of a new 
genus and the reference of this to the family Paractidse is necessary. 

Synonyms. — jcfiwiflt reticulata. 一- Daua U. S. Expl. Exped., 1846. 

Metridium reticvlatum. 一 Milne-Edwards, 1857. Verrill, 1868. 

jc<i»ioZo&arertcMZa《fl. — Gosse, 1860. Antholoha reticulata, ― R. Hertwig, 1882. 
No8. 737, 738. Station : Port Otway, Patagouia. Littoral. Two specimens. 
No. 739. Station : Lota, Chile. Littoral. One specimen. 

No. 740. Station : Charles Island, Galapagos Archipelago. Littoral. One specimen. 

I have very little to add to tlie description Hertwig has given of this | 
form. 1 do not find, however, that the margin of the disc is "swollen 
like a pad," but on the cootrary the uDpermost portion of the column 

Genus ANTHOLOBA, Hertwig. 

15. Antholoba reticulata, (Dana) Hert. 

VOL. XVI, 1 
1893. J 



wall is in some specimens thinner than it is farther down. The presence 
of a pad may be due to contraction. 

The sphincter muscle, as Hertwig pointed out, extends from the up- 
per to the lower end of the wall. Its shape may be of generic impor- 
tance, since it does not present the sudden widening near the margin 
which is to be seen in the forms I have referred to the genus Paractis, 
but tapers off very gradually indeed as it passes down the column. 

The specimens I examined did not possess reproductive organs, so 
that I can not decide the question Hertwig has raised regarding the 
hermaphroditism of this form. 

ParactidsB with thick column wall ; margin lobed; tentacles short, 
situated near the margin, tbe mesoglcea thickened toward their bases, 
80 as to give them a more or less bulbous appearance. Sphincter 
muscle rather weak (sometimes absent!). 

The genus Aetinernus was established by Verrill (79) for a deep-sea 
form obtained off the more northern portion of the east coast of the 
United States. Verrill's definition and description speak of the margin 
below the tentacles being "divided into acute lobes or teeth continuous 
with the body wall," the tentacles being adnata to these teeth. This is 
the appearance which Aetinernus nob His presents, but I have preferred 
to speak of the teeth as thickenings of the inesogloea of the bases of the 
tentacles, since this more nearly describes what obtains in A.plebeius, 
and probably also in A. mginatm. The sphincter muscle is quite weak 
in A.plebetm, as will be seen from the following description, and ap- 
parently is wanting in A. nobilis, being indistiuguishable with a pow- 
erful lens. This character offers a marked difference, independent of 
the nature of the tentacles between this genus and Antholoba. 

The similarity which the figure of Polysiphonia tuberosa given by 
Hertwig ('82) shows to an Aetinernus is very striking and suggests 
its possible reference to the latter genus. The lobed margin, the ba- 
sally swollen tentacles, the disc marked with radiating grooves, the 
chalice-like shape of the column, are all similarities which attract at- 
tention. The sphincter muscle, too, though differing in shape from 
that of A.plebeiusy to be described below, is nevertheless mesodermal 
and by no means powerful. The principal characteristic upon which 
Hertwig relies in the establishment of the geims is found in the rather 
large openings at the tips of the tentacles. Such openings are known 
to be of frequent occurrence, and their enlargement within certain limits, 
unaccompanied by a marked abbreviation or other alteration of the 
tentacles, can not be considered sufficiently distinctive for the forma- 
tion of a new genus. It seems to me that a reference of Polysiphonia 
tuberosa to Verrill's genua Actinernm will place it with forms to which 
it is far more closely related than it is to Polystomidium. (See Appen- 
dix, p. 209.) 

Genus Aetinernus, Verrill. 


16. Actinemus plebeius, sp. nov. 

Plate XXIV, Figs. 42-45. 

No. 711. Station 2791. Lat. 38。 08' S.; long. 75° 53' W. Depth, 677 fathoms. 
One specimen. 、 

The body is calyciforni and measures about 5徵 in height, with a di- 
ameter of about at the disc. The base on the other hand measures 
only 2.5 復 in diameter. It was probably adherent, thoiigU from its 
great distortion in the single specimen it is difficult to be certain what 
its character may have been. 

The column wall is rather soft in consistency, though relatively thick, 
and its surface being somewhat torn into threao^ has a ratUer ragged 
appearance. The ectoderm is almost entirely macerated away, but the 
few fragments that remain show that it was of a chocolate brown color. 
The sphincter is embedded in tUe mesoglcea, not far from its endodermal 
surface. It extends some distance down the column wall, but is very 
narrow. In sections (PI. xxiv, Fig. 43) it is seen to consist of a series of 
cavities placed one above the other, for the most part in a single row, 
each cavity being separated from its neighbor by a distinct partition 
of mesoglcea. Each cavity is occupied by a mesoglceal network of 
fine fibre, in the circular or oval interstices of which the muscle cells are 

The margin is tentaculate and wavy or lobed in outline. The tenta- 
cles are about ninety-six in number and are arranged in two or three 
cycles at the margin. They are of a purplish -brown color and are 
short and slender, each being provided at the outer surface of its 
base with a marked mesoglceal thickening (PI. xxiv, Fig. 42), which ex- 
tends a short distance upwards towards the tip upon the outer surface 
of the tentacle. The longitudinal muscles are weak and are not em- 
bedded in the mesogloea. 

The disc is concave and of a wine-purple color and is marked with 
radiating ridges, due to the roofs of the inter- and intra-meseuterial 
spaces being pouched out. The radiating muscles are ectodermal and 
not at all embedded in the mesogloea. 

The mouth forms an elevation at the center of the disc. It is provided 
with two well-marked sipbonoglyphes. Tlie stomatodaeuin is longitud- 
inally ridged, the walls of tlie deep siphonoglypbevS >eiiig on the other 
hand smooth. The ectoderm of the stomatodsBum is of a deep wine- 
purple color. 

The mesenteries are arranged m four cycles, though indications of a 
fifth and sixth cycle were present, neither of them being, however, per- 
fect. In a sextant of the wall examined only one pair of mesenteries of 
the sixth cycle was present, and five pairs, instead of eight, of the fifth 
cycle. Only the six pairs of mesenteries of the first cycle are perfect, 
and only the mesenteries of the third and fourth cycles are gonophoric. 
The musculature, both longitudinal and parieto-basilar, is very weak. 
What corresponds to the muscle pennon is very low, the mesogloea being 
raised into short, blunt processes which carry the muscle cells and give 



to the surface of the mesentery on which they occur a crenate appear- 
ance in transverse sections (PI. xxiv, Fig. 44). The endoderm of tbe 
mesenteries and that of the body wall is of a purplish-brown color, paler 
than tbe stomatodaeum, while tbe mesenterial filaments, in whole or in 
part, have the same deep wine color which has been described for stoma- 
todaeum and disc. The mesogloea of the reproductive region of the gono- 
phoric mesenteries is much thickened, as is shown in PL xxiv, Fig. 45. 

Genus ACTINOSTOLA, Verrill. 

Para^tidae usually of large size, with firm, leathery wall, which may 
be somewhat corrugated or folded, but is not furnished with verrucae. 
The margrn is not lobed and is tentaculate; the tentacles are short and 
stout, fluted and with their loDgitudiual musculature embedded in the 
mesogloea. Sphincter well developed, extending a considerable dis- 
tance down the column wall and not expanding abruptly above. 

The genus Actinostola was established by Verrill (,83) for a species 
which he had previously ('82) described as Urticina callosa. In his de- 
scription of the genus he states that the column is "covered with large, 
irregular tubercles not having the power of adhering to foreigu substan- 
ces," and in the description of the species ('83) he says : "The surface 
of the colamn is usually more or leas covered with low, irregular, often 
flattish verrncae, which become more and more prominent and -some- 
times form Ion gitndinalfseries or crests on the upper part, but facie out 
to mere wrinkles toward the base." In specimens of A. callom, wliich 
I have, through the kindness of Mr. Eathbuii, been able to examine, I 
could find nothing that could properly termed verrucje, or even 
tubercles, though the surface of the column wall was more or less cor- 
rugated, resembling in some specimens beaten silver, and bore irregular 
ridges of mesogloea near the margin. The Albatross specimens present 
the same appearance, though in one case tlie corrugations are sufficiently 
strong to give an almost warty appearance to the column. 

Verrill considers the genus Actinostola to be allied to Bolocera, Ur- 
ticina, and especially to Aetinauge. What tbe genns Urticina, may 
embrace remains to be seen; but the other two genera mentioned have 
certainly only very remote affinities with Actinostola^ Bolocera being 
related to the Antheadae, and Aetinauge one of the Sagartid genera. 

17. Actinostola callosa, Verrill. 

Plate XXIV, Fig. 46; Plate xxv, Figs. 47-52. 

Synonym : —— Urticina callosa, Verrill. 1882. 
No8. 714-715. Station 2792. Lat. 0^ 37' S. ; long. 81^ 00' W. Depth, 401 fathoms. 
Four Bpecimeus. 

No. 721. Station 2807. Lat. (F 24' S. ; long. 87^ 06' W. Depth, 812 fathoms. Two 

No. 723. Station 2818. Lat. 0。 29' S. ; long. 89。 54' 30" W. Depth, 392 fathoms. 
One specimen. , 

The Albatross specimens denoted above I can not distiuguisli from 
Verrill's A. callom, with authentic specimens of which I have carefully 

Digitized by 


compared tliem. They measure about 8。*» in height, with a diameter of 
5.5c™. Most of the specimens (PI. xxv, Fig. 47) are only partially con- 
tracted, allowing the tentacles to partially protrude, but in some they 
are entirely concealed from view. 

The base is flat, marked with fine radiating lines, and has the lim- 
bus folded back over its edges in all the specimens. The column is 
nearly cylindrical, and slightly smaller above than below. Its wall lias 
a firm, parchment-like consistency, and is variously corrugated, in part 
owing to contraction. In tlie more fully expanded specimens the sur- 
face has somewhat the appearance which beaten silver or other soft 
metal presents, while in others the corrugations may be sufficiently 
pronounced as almost to justify the designation of irregular tubercles. 
There are, however, 110 indications of verrucse. . Below the margin the 
mesogloea is rougher than elsewhere, and is raised into irregular ridges. 
The column wall has a snowy white appearance, the ectoderm in all 
the specimens having almost disappeared ; the fragments of it whicli 
remain in some specimens seem to indicate that it was of a pale, brown- 
ish-purple color. The sphincter (PI. xxv, Fig. 51) is well developed 
and extends a considerable distance down the column wall. In its 
upper part it does not occupy the entire width of the column wall, but 
lies throughout its course nearer the endoderinal surface than the ecto- 
dermal, its cavities passing, in fact, directly into the ordinary circular 
musculature of the endoderm. It does not expand suddenly above, 
but its upper part, tUough larger than the middle region, tapers off 
very gradually as it is traced downwards. In its upper part the closely 
packed muscle cavities sliow a tendency to be arranged in longitudinal 
bands (PI. xxv, Fig. 52) separated from one another by streaks of 
nearly homogeneous .mesogloea, and recalling the arrangement which 
Hertwig ('82) has described for Ms Bysactis crassicornis. 

There is no well-defined margin, the tentacles being inserted upon it. 
They are rather numerous, situated close to the margin, and are short 
and stout, with well-marked pores at their extremities. They have a 
more or less decided pink or salmon color, and are rather indistinctly 
longitudinally fluted. Their longitudinal' musculature is imbedded in 
the rather thick mesogloea (PI. xxv, Fig. 48), as is also the radial mus- 
culature of the disc. This portion of the body is smooth and concave 
and has the same pinkish color whicli occurs in the tentacles. The 
moutU is wide, and the stomatodaeum is about half the length of the 
body. It is longitudinally ridged, and has two well-marked, deep 
siphonoglypbes with smooth walls, which are continued down below 
the lower edge of the stomatodseum, almost to the base. 

Twenty-four pairs of mesenteries reach the stomatodseum, but twelve 
of them are united to the stomatodaeum to a less extent than the other 
twelve. In adcjition to these there is another cycle of twenty-four im- 
perfect pairs, whicli may be counted as the fourth cycle, while the fifth 
cycle of forty-eight pairs, also imperfect, presents the anomalous con- 



dition of. one mesentery of each pair being much more highly devel- 
oped than its fellow (PI. xxv, Fig. 46). One of each pair is quite small, 
without reproductive organs aud mesenterial filaments, aud hardly 
projects above the column endodferm, while its fellow is fairly broad, 
and carries reproductive organs and a mesenterial filament. A similar 
disparity, though less marked, is to be found in the pairs of the 
fourth cycle, but I could not distinguish it in the third cycle. The re- 
lation of the small to the large mesentery of each of the unequal pairs 
seems to be constant, and is shown in the diagrammatic figure (PL xxiv, 
Fig. 46). It will then be seen that in the fiftU cycle (v) the small mesen- 
teries are those nearest the mesenteries of the fourth cycle (iv), while in 
the fourth cycle the strongest mesenteries are those nearest Ihe pairs 
of the first and second cycle. A few irregularly disposed mesenteries 
of the sixth cycle could also be seen. The mesenteries of the fourth 
and fifth cyles are gonophoric. 

As regards the musculature of the mesenteries, it is not very 
strongly developed. At the base of each mesentery (PI. xxv, Fig. 50) 
there is a strong development of muscle processes on both sides, pro- 
ducing a basal muscle (bm) similar to wbat occurs in the Edwardsiae, 
and to a less extent in many Hexactinians. In the mesogloea of the 
basal region of the mesenteries of the first three cycles some cavities 
are to be observed similar to, but less highly developed, than those al- 
ready described for Bolocera oecidua, and like those developed in connec- 
tion with the parieto-basilar muscle {pbm)^ which forms a slight pro- 
jection oil one side of the base of the mesenteries. The longitudinal 
muscles cover all the muscular portion of the mesenteries in an almost 
uniform layer, only toward the inner edge of the muscular regiou be- 
coming longer aud forming a rather weak muscle pennon (PI. xxv, Fig. 
49). The muscle processes, especially in the pennon, show a tendency 
to be arranged iii groups ou more or less distinct blunt processes of 

Amongst the Challenger material Dysactis crassicornis presents cer- 
tain features of marked similarity to Actinostola callosa. The general 
arrangement of the muscle cavities of the sphincter muscle seems to be 
identical in the two forms, and the peculiar arrangement of the 
mesenteries of the younger cycles shows interesting similarities. 
There are, however, certain differences in the arrangement, which 
have made me hesitate to identify the two forms, tUough I am inclined 
to believe that Dysactis crassicornis is to be properly referred to the 
genus Actinostola, and that it is even probable that it may be identical 
with A. callosa. There can be little question that its reference to 
Milne-Edwards' genus Dysactis is incorrect, since we know that two 
at least of the forms referred by its author to it, D. annulata (Lesueu?:') 
and !)• biserialis (= Aiptasia conchii Gosse), are Sagartids, while 
D. chilemis is also referred to that family by Verrill and Andres. If, 
therefore, the forms referred to Milne-Edwards' genus are Sagartids it 

Digitized by 


can scarcely be proper to associate with them Paractids. In cases 
like this where the definition is imperfect we have to interpret the 
genus from the forms which have been assigned to it and not vice versdj 
and a more perfect definition of the genus Dysactis will include a men- 
tion of the occurrence of acontia and ciuclides. (See Appendix p. 209.) 

18. Actinostola ezcelsa, sp. no v. 

Plate XXVI, Figs 53-56. 

No. 696. Station 2770. Lat. 48° 37' S. ; long. 65c 46' W. Depth, 58 fathoms. One 

No. 698. Station 2771. Lat. 51。 34' S. ; long. 68。 00' W. Depth,, 50^ fathoms. Two 

This very striking form (PI. xxvi, Fig. 53) measures about 6 cm in 
height and from 5.5 to G"" in diameter. The base is evidently 
adherent and the limbus is not folded over it, as was the case in A. 

The column is cylindrical, narrowing slightly towards the margin, and 
is apparently capable of little contraction. Its walls are firm, and for 
the most part smooth, though in contracted specimens irregular longi- 
tudinal ridges are to be seen below the margin ; these, however, seem 
to be due to the state of contraction. The ectoderm of the column ha.s 
a pale brown or buff color; where it has been macerated away the sub- 
jacent mesogloea is seen to be cream white. The sphincter muscle 
(PI. XXVI, Fig. 54) is fairly strong, but nevertheless is unable to over- 
come the resistance offered by the firmness of the column mesogloea, so 
that in none of the specimens are the tentacles concealed from view. 
In shape the sphincter differs markedly from that of A. callosa. It 
forms a delicate network, occupying almost the entire thickness of the 
mesogloea in its upper half, and its inner surface passes into the gen- 
eral circular musculature of the column wall. There is no tendency 
for the muscle cavities to arrange themselves in longitudinal rows as in 
A, callom, but rather in horizontal lines perpendicular to tbe surface of 
the column. The column wall is less thick in its uppermost part than 
a little lower down, and consequently the thickest portion of the sphinc- 
ter is below its uppermost edge, in fact almost half-way down. In its 
lower part it is thin, lying* close to the endodermal surface of the meso- 
gloea, and is -prolonged downwards some distance in this condition, 
gradually becoming lost in the muscle processes of the circular muscu- 
lature of the column wall. 

There is no definite margin, the tentacles occurring at the junction of 
the disc and column wall. They are rather numerous, numbering per- 
haps one hundred and ninety- two, and are short and stout, with pores at 
their extremities. They are longitudinally fluted, and also transversly 
grooved, so that the surface seems much corrugated. The mesogloea 
of the tentacles is almost entirely occupied by the longitudinal muscles 

Digitized by 



( PI. XXVI, Fig. 56) ; ill the elevations which give rise to the flutiiigs, how- 
ever, it has a very delicate structure resembliug greatly typical areolar 
tissue with its connective.tissue corpuscles. 

The month is large; the stomatodaeum is irregularly ridged longi- 
tudinally, and the siplionoglyphes are deep and prolonged below the 
lower level of the stomatodaeum. 

The mesenteries are arranged in ninety-six pairs, of which only 
those of the first two cycles, twelve in all, are perfect. These, together 
with the mesenteries of the third cycle are sterile, the reproductive 
organs occurring only on the mesenteries of the fourth and fifth cycles. 
The longitudinal musculature (PI. xxvi, Fig. 55) is fairly strong but does 
not form any distinct pennon upon the surface of the mesentery. The 
muscle processes show a tendency, especially in the basal portion of the 
mesentery, to be grouped upon low elevations of the general mesogloea. 
The parieto-basilar muscle (pbm) is well developed and forms a decided 
projection upon the basal portion of the mesenteries, which portion, 
where the parieto-basilar occurs, contains a number of cavities, evi- 
dently developed, as in B. occidua, in connection with the growth of the 

19. Aotdnostola pergamentacea, sp. nov. 

Plate XXVI, Figs. 57 and 58; Plate xxvii, Figs. 59-63. 

No. 695. Station 2769. Lat. 45。 22' S. ; long. 64。 20' W. Depth 51* fathoms. Five 

These specimens (PI. xxvi, Fig. 57), which seem to belong to the genus 
Actinostola, are very much macerated, the tentacles having dissolved 
into shreds, so that it is impossible to ascertain their shape or struc- 
ture. The specimens measure 3^"、 in height and 2**"^ in diameter. 

The base is evidently adherent and larger in diameter than the 
column. This is almost cylindrical, enlarging somewhat at the margin 
and limbus. Its walls are smooth, firm, and parchment-like, being 
brittle rather than tougli, and readily broken. It is pure white in 
color, the ectoderm, however, being entirely absent. The sphincter 
PI. XXVII, Fig. 59) resembles in general appearance that of A. exeelsay but 
is by no means as strong. None of the specimens show the slightest 
trace of the margin being infolded over the tentacles, and this is not 
remarkable, considering the stiffness of the column mesogloea. 

The tentacles seem to have been numerous, perhaps one hunared 
and ninety-two, though this is merely an estimate, since they are too 
badly macerated to allow of a count. Their longitudinal musculature is 
imbedded in the mesogloea in a number of small cavities (PI. xxvii, Fig. 
60). The disc is roughened by radiating rows of small tubercle-like 
elevations, and the radial musculature resembles that of the tentacles, 
though in one specimen the cavities were elongated and separated by 
narrow trabeculae of mesogloea, presenting the appearance shown in 
PL XXVI, Fig. 58. 

Digitized by 


The inoutU is prominent. The siphonoglyphes are deep and long^er 
than the stomatodaBum. All the mesenteries, with the exception of 
the youngest cycle, are perfect; there are apparently five cycles, tlie 
mesenteries of the third and fourth cycles being gonophoric. The 
muscle processes of the longitudinal muscles are developed over the 
entire muscle-bearing surface of the mesentery, increasing slightly 
towards the inner edge of this surface to form a weak peunon. In the 
upper part of the mesenteries, above the region where the parieto- 
basilar occurs the parietal part of the mesentery is somewhat thick- 
ened, and the muscle processes in this thickened region are somewhat 
more numeroas and more slender than elsewhere (PI. xxvii, Figs. 62-63). 
Over the general surface of the mesenteries the processes are com- 
paratively stout (PI. XXVII, Fig. 61). The parieto-basilar muscle pre- 
sents essentially the same characteristics as in A, callom, the mesogloea 
ill the region occupied by it baving small cavities enclosed iu it. As in 
A, callom also a basal muscle is present (PL xxvii, Fig. 62), but it has 
relatively but a slight development. 

Paractidae of moderate size, with thick, though rather soft, column 
wall; no tubercles or vermcae, though the upper portion of the coliimu 
is marked by more or letss distinct longitudinal ridges running to the 
bases of the tentacles. Margin teiitaculate, not lobed ; tentacles short, 
but slender, not swollen at the base. Sphincter muscle rather weak, 
lying close to the endoderm. 

I have established this genus for the reception of a form which does 
not seem to be assignable to any of the genera of Paractidae as they 
are here understood. Tlie weak sphincter and slender tentacles ex- 
clude it from the geDua Actinostola; the absence of a marked aiJata- 
tioii of the sphincter aud the occurrence of ridges upou the upper part 
of the column, running to the bases of the tentacles, show it to be dis- 
tinct from the genus Paractis. The ridges are hollow, with rather del- 
icate walls, and resemble those found in certain Sagartids which pos- 
sess a capitulum. The absence of acontia, however, precludes the 
association of the form about to be described with the Sagartidae. 

No. 728. Station 2839. Lat. 38" 08' N. ; long. 118。 40' W. Depth, 414 fathoms. 
Fourteen specimens. 

The largest specimens (PI. xxvii, Fig. 64), measure 2.5«™ in height, and 
3i3cm in diameter. All are contracted, the tentacles and upper por- 
tion of the column being infolded. The alcohol in which they are pre- 
served is stained a very distinct yellow, and when specimens are placed 
in fresh alcohol this quickly assumes the same coloration. The pig- 

Genus P YCNANTHUS, gen. nov. 

20. Fycnanthus maliformia, sp. nov. 

PI. XXVII, Figs. 64-67; PI. xxviii, Fig. 68. 


ment seems to saturate the alcohol quickly, fresh alcohol continuing to 
extract more of it even after several chaugings. 

The base is thiu, allowing the mesenteries to be seen through, when 
the more or less membranous brown coating which covers it is re- 
moved. The margin of the base in all the specimens is concealed by 
the limbus being reflected over it. 

The colamii is white, the ectoderm having been entirely removed, 
and is irregularly corrugated ; no tubercles or verrucje are present, 
however. The mesoglcea is very thick, measuring iu one specimen at a 
point a short distance above tlio limbus as much as 3.5 陣 in thickness. 
It is not, however, harsh or resistant to the touch, but on the contrary 
is rather soft, and in structure is almost homogeneous or hyaline, with 
small cells scattered through it. Towards its upper part are a number 
of ridges, which are hollow and thin-walled, and pass to the bases of 
the tentacles of the outer row. The sphincter muscle extends a con- 
siderable distance down the column wall (PI. xxvii, Fig. 65), but is 
throughout thin. It lies throughout its entire extent close to the eudo- 
dermal surface of the mesogloea, passing into the circular musculature 
of the column wall. Above it is very slightly thickened, but not at all 
as in Paraetis. The muscle cavities for the most part show little ten- 
dency towards any regular arrangement (PI. xxvii, Fig. 66), though 
towards the lower edge of the muscle tlioy are somewhat elongated, and 
arranged in lines nearly perpendicular to the surface of the inesoglcea. 

The margin is tentaculate. The tentaoJes are arranged iu four cycles, 
12, 12, 24, 48. The ridges upon the upper surface of the column run to 
the bases of the outer tentacles, and from the bases of the inner ones 
ridges extend outward, but only for a short distance, losing themselves 
before they reach the outermost cycle of tentacles. The mesogloea of 
the bases of the tentacles is only very slightly thickened, and tbe longi- 
tudinal muscles of tbe tentacles are imbedded in the mesoglcea. In the 
disc the radial muscles are mesogloeal and are arranged iu a very char- 
acteristic maimer (PI. XXVIII, Fig. 68), recalling what. Hertwig has figured 
for Dysactis crasaicornis. Opposite the insertions of the mesenteries 
into the disc the radial musculature is interrupted so that it is di- 
vided into radial bands, each separated from its neighbors by a depres- 
sion on the surface of tbe disk. Each radial baud appears to be a 
single flattened cavity, traversed by perpendicular, somewhat branch- 
ing, fine trabecular of mesogloea, which divide the large cavity into a 
great number of smaller ones, in which lie the muscle cells. 

The stomatodae ひ m is longitudinally ridged and is continued down- 
wards almost to the base. Tbe siphouoglyphes are deep, and near 
their lower extremities two transverse folds, lying one above the other, 
project across the cavity of each, closing it below. 

The mesenteries are arranged iu ninety-six pairs, the youngest cycle 
of forty- eight pairs being indistinguishable to the naked eye. The 
twelve pairs of the first two cycles are perfect^ the twelve ter お aries 

Digitized by 


also reaching the stoinatodaeum, but being united to it to a less extent 
than are the primaries and secondaries. The reproductive organs are 
borne upon the mesenteries of the tbird and fourth cycles. The lon- 
gitudinal muscles of the mesenteries do uot form a distinct pennon, 
(PL XXVII, Fig. 67). In the perfect mesenteries the processes Which sup- 
port the muscle cells arise in bunches from stout elevations of the 
mesogloea. The parieto-basilar muscles extend ouly a very short dis- 
tance up from the base, and in sections through the middle of the 
column are not to be distinguished. No acoutia appear to be present. 
The endoderm is considerably macerated, so that the form of these 
structures, if they existed, could not be made out 5 I base my state- 
ment as to their absence in the absence of netnatocysts in the tissues 
lying ill the body cavity, the macerated remains of the mesenterial 

Paractidse of moderate size, crateriform in shape, with the mesoglcBa 
of the column wall rather thick but soft; surface of column rugose in 
contracted forms, but without verrucas or warts; no capitulum with 
longitudinal ridges. Sphincter muscle relatively weak, lying close to 
the endoderm; margin not lobed, teiitaculate; tentacles numerous, 
situated close to the margin, short, slender, not bulbous at the base. 

The form for which I establish this geuus approaches somewhat m 
appearance an Actinernusj liaVing the short tentacleii concentrated near 
the margin as in that genus, a large portion of the disc being left un- 
covered. The absence, however, of any bulbous enlargements or thick- 
ening of the mesogloea at the bases of the tentacles induces me to place 
the form in a separate geims, which, from the cup-shaped form of the 
specimens to bo referred to it, I name Cymbactis {Kv/u/Srj ~~ a drinking 

Plate XXVIII, Figs. 69-71. 

No. 732. Station 2839. La に 33。 08' N. ; long. 118。 40' W. Depth, 414 fathoms. Six 

All the specimens seem to be immature, as I did not succeed in flnd- 
ing reproductive cells in those I examined. The largest specimen 
measured 2 cm in height, with a diameter at the margin of 2.5 "", and 
at the base of 1.3 。". 

The base is adherent. The column which gradually enlarges from the 
base to the margiu, producing a more or less cup or vase shaped form, 
(PI. XXVIII, Fig. 69), is rather thick-walled, but soft to the touch, the 
mesogloea not beiDg of fibrous structure but hyaline. The ectoderm 
has macerated away from all the specimens, but when a trace 01 it is 

Genus CYMBACTIS, gen. nov. 

21. Cymbactis faeculenta, sp. nov. 

VOL. XVI, , 

1893. J 



left it may be seen to be of a chocolate brown color. lu consequence of 
the absence of ectoderm the column " white, tliough in some of the 
smaller specimens, in which the mesogloea is thinner, the color was a dark 
slate blue, due to the dark pigment of the endoderm showing through. 
The column wall is very rugose, probably due to contraetiou, and »hows 
no signs of possession of verrueae or perioaueiit warts. The sphincter 
(PI. xxviu, Fig. 70) is weak, compared to what it is in most Paractids, 
and is for the most part confined to a thm layer immediately external 
to the eiidoderm. Toward its upper part a few scattered and isolated 
cavities are to be seen deeply imbedded iii the luesoglcea, apparently 
undergoing degeneration. 

In the contracted specimens the tentacles are concealed partially by 
ail infoldiug of the margin, but this infolding is not carried far enough 
to conceal the disc and the wide mouth. The tentacles are situated 
close to the margin in about five cycles, and are apparently about nine- 
ty-six: in number. They are short, acumiuate, and slender. Their ec- 
toderm and that of the disc seems to be of the saQio color as that of 
the column. The radial musculature of the disc and the longitudinal 
muscles of the tentacles are imbedded iu the mesogloea. The mouth is 
wide and leads into a stomatodi^iim which reaches nearly to the base. 
The siphonoglyphes are well developed. 

The stomatodseal ectoderm and the eiidoderm throughout is of a dark 
wine color; the pigment occurs in the form of granules scattered 
through the cells, and is insoluble iu alcohol, turpentine, and xylol. 
The mesenteries appear to number twenty-four pairs, half of which are 
perfect. Their mesogloea is thick, and there is no special muscle pennon, 
the longitudinal muscles being comparatively weak (PL xxviii, Fig. 
71). No reproductive organs could be made out. 

Actininae with sphincter muscle imbedded in the mesogloea, usu- . 
ally with only a few perfect mesenteries; furnished with acoutia. 

According to the above defluition the Sagartid^e will form a group 
parallel to the Paractidse, and distinguished from them by the presence 
of acontia. Whether this is a character of sufficient importance 
for a family diagnosis and indicates phyletic affinity of all the forms 
which present it future observation must determine. It seems at present 
convenient to associate all Actiniuae with acontia iu a single family, 
though it may be necessary to recognize iu the family various sub- 
families, as several authors have already done. Haddou ('89) has 
discussed the Hmitatious of the family as tliey have been placed by 
various authors, and accordingly it will be unnecessary to repeat such 
a discussion here. The same author lias established a new subfamily 
Chondractininne, which may, for the present, be adopted, though it 
seems not improbable that it is practically identical with the sub- 

Family SAGARTII)^. 


family, Phellinae, which was separated from the SagartiddB by Verrill 
('67)^and recognized by Andres (,83) and Hertwig ('88). I have in a pre- 
vious paper (1889) proposed the separation of the SagartidsB into the 
subfamilies Sagartiiiae and Phellinae, but since Haddon's subfamily is 
somewhat more extensive than and probably may include tbe genus 
Phellia, it seems advisable to adopt it. 

Sagartidae with the ectoderm naked, the acoDtia being emitted from 
the mouth and through the column wall, in which definite openings 
(cinclidevs) are present (always?) for their emission. 

S&gartinae with smooth column destitute of verrucsB and with no 
special arrangement of the cinclides; margin tentaculate; tentacles 
concealed in contraction, the sphincter being fairly strong. 

In alcoholic specimens it is not always possible to be certain as to 
the arrangement of cinclides, and some of the forms which I assign to 
this genus may possibly be more properly referable to some other Sa- 
gartian genus. The absence of verrucae, the tentaculate margin, and 
the concealment of the tentacles in contraction are points which assist 
ill determining the assignment of a form to this genus. 

Nos. 710-956 Station 2785. Lat. 48。 09' S. ; long. 74^ 36' W. Depth, 449 fathoms. 
Numerous specimens. 

The specimens were adherent to a dead coral, ami were for the most 
part strongly contracted, forming a low rounded cone with a widely 
expanded base (PI. xxviii, Fig. 72). In these the tentacles were com- 
pletely concealed, but in a few forms the contraction was not so great, 
and the tmtacles were partly visible. Such specimens measured from 
1.1 to 1.3CU1 lieiglit, with a diameter at the upper part of the column of 
about 1.1 噴 and at the base of about 1.5 or 2 ゆ'. 

The base is provided with a brown membranous covering, evidently 
a secretion of its ectoderm cells. The ectoderm in all the specimens 
has been entirely macerated away from the inesogloea of the column 
wall, which has a milky white color. It is tolerably firm and parch- 
ment like, though not very thick, and is for the most part smooth, 
though ip some specimens more or less wrinkled by contraction. In 
the upper part of the column delicate longitudinal ridges can be seen, 
which become stronger as they approach the margin and recall the 
capitular ridges of Actinauge; they are not, however, visible in the less 
contracted specimens, and seem to be produced by the contraction of 
the spbincterj aod to be due to a certain extent to the pergamentaceoua 

Subfamily Sagartin^. 


22. Sagartia lactea, sp. no v. 

Plate XXVIII, Figs. 72-75; Plate xxix, Fig. 7H. 

▼OL. XVI, "I 

1893. J 



consistency of the mesoglcea. Tbe sphincter muscle (PI. xxviii, Fig. 
73) is fairly strong and in its upper part occupies nearly the whole thick- 
ness of the mesoglcea, being separated from the ectoderm and endoderm 
by thin layers of niesogloea. It is composed of very numerous more or 
less circular (in section) cavities lined with muscle cells, and so closely 
arranged as to be separated only by very narrow bands of fibrous me- 
soglcea (PI, XXIX, Fig. 76). In consequence of their arrangement this 
portion of the column wall, under low magnification seems to have a 
reticular structure. The sphincter extends a considerable distance 
down the columa, becoming thinner and having the cavities more sep- 
arated in its lower part, until finally they are scattered singly or in 
pairs ill the lowermost portions. 

The tentacles are slender and acuminate, and their number I estimate 
at slightly below one hundred, though I was unable to make adefiuite 
count. They have a cream-white color. Their longitudinal muscula- 
ture is ectodermal, and the inesoglceal supporting processes are fairly 
strong. Large nnmbers of iiematocysts occur in their ectoderm. 

The disc bas strong radiating ridges correapondi ug to tbe endoccels 
of the first and second cycles of mesenteries, and has its radiatiug 
musculature ectodermal, like the longitudinal muscles of the tentacles. 
【11 the ectoderm of the disk are numerous oval or spherical bodies, of 
a granular structure, which stain deeply with borax carmine. I could 
not detect a nucleus in any of them. Their abundance and general ap- 
pearance seem to preclude the idea that tbey are foreign bodies, and 
the only explanation as to their significance which suggests itself is 
that they are glandular bodies. The preservation of the ectoderm was 
not sufficiently perfect, however, to allow of any certainty ou this point. 

The stomatoda^um is rather small In diameter, and possesses about 
ten loDgitiidiual ridges; in some specimens there was only a single 
siphonoglypbe, but whether this is a characteristic arrangement I can 
not say. Judging from tbe observations of G. F. and A. Y. Dixon 
on various species of Sagartia ('88) and my own ('91) ou Metridium 
marginatum J it is more probable that there is a variation in the number 
of siphonoglyphes, some specimens possessing only one and others two. 
As ill the case of Metridium and Sagartia I'emtsta, there is only one pair 
of directives in those specimeus of 8. lactea which possess a single 

The mesenteries are arranged upon the decamerous plan, there being 
in all ten pairs of perfect mesenteries, all ot* which, with the exception 
of the directives, are gonophoric. I was iu hopes that it might be possi- 
ble, from the distribution of the reproductive organs upon the mesen- 
teries, to ascertain which of the mesenteries of the second cycle it was 
which had failed to develop, tlie normal liexamerous arrangement 
being thus converted into a decamerous one; but in this I was disap- 
pointed. Counting the ten perfect pairs of mesenteries as representing 
two cycles, one of which, the second, is not quite complete, there is present 

Proc. M. 93 —— 12 


a third cycle of ten pairs, all gonophoric, a fourth one of twenty pairs 
destitute of reproductive organs, and indications in some exocoels of a 
fifth pair, which is, however, incomplete. The mesenteries are thin, and 
their musculature not very markedly developed (PI. xxviii, Fig. 74). 
The acontia are fairly numerous and show a large development of gland 
cells (PI. XXVIII, Fig. 75). The convex surface of an acontium is occupied 
mainly by Dematoeysts, between which a few scattered coarsely granu- 
lar gland cells occur, while immediately below the nematocysts these 
cells are very abundant, as they likewise are at the sides and towards 
the concave surface. For the most part they staiu deeply with borax 
carmine, though many 一 probably those in which tbe glandular products 
are more completely elaborated ― refuse to take the stain and show a 
yellow color. In one specimen I found the acontia protruding from tbe 
mouth, but could not find any emitted through the column wall, although 
in sections through the wall fine canals can be readily observed which 
have no appearance of being artefacts, and probably are cinclidal. I 
could discover no definite arrangement of these canals. 

There are three interesting features about this Sagartid : (1) Its 
decamerisiii. There are ten pairs of perfect mesenteries, and the imper- 
fect mesenteries are arranged symmetrically to these ten, those of the 
next subordinate cycle developing in the exocoels between adjacent 
pairs ofpecf^ct mesenteries. I Lave already suggested ('91) that this con- 
dition probably arises by the suppression of a pair of mesenteries of the 
typical second cycle, so that this cycle consists of four pairs only in- 
stead of six. Whether or not it is the same pair that is suppressed in 
each case in wliicli decamerisDi occurs can not be stated at present. In 
the decamerous Haleampids it has been seen that it is the mesenteries 
on either side of the sulcular directives that have disappeared, but it is 
not impossible that in sporadic cases of decanierism, such as we have in 
S. lactea, that it is the mesenteries on either side of the sulcar directives 
that have disappeared, or even the lateral mesenteries of the second 
cycle. However that may be, it is certain, I think, that we must re- 
gard the ten perfect mesenteries of a decamerous form as equivalent to 
the first and second cycles of a hexaineroiis form. It follows from this 
(2) that we have in S, lactea another instance of a Sagartid iu which 
more than the six primary mesenteries are perfect. Hertwig ('82) as- 
sumed as a character of his family Sagartidse tlie presence of only six 
perfect mesenteries, which were also sterile, but von Heider ('77) had 
already shown that there were numerous perfect mesenteries in Cereus 
pedu7iciilatu8, and F. Dixon ('88) has since shown that in those Sagar- 
tias which Gosse considered typical species of the genus there are more 
than six pairs of perfect mesenteries. It is certainly a fact that the 
majority of Sagartid 8 whose anatomy we know possess only six pairs 
of perfect mesenteries, but too many exceptions exist for this peculiar- 
ity to be included in the definition of tbe genus. But not only does 8. 
lactea have the mesenteries of the second cycle perfect, but (3) the 

VOL. XVI, 1 
1893. J 



mesenteries of the first cycle, with the exception of the directives, are 
not sterile. Here again we have a feature which places this form out- 
side the pale of Hertwig's genus SagarUa, but it shares this distinction 
together with Aiptasia sp! and Aiptdsia palliday whose peculiarities in 
this respect I have already pointed out ('89a). 

I have considered the form described in the following pages to be 
the representative of a new species. I do so, however, with considera- 
ble hestiation. Several Sagartids have been described from the west 
coast of America, by Lesson (,30), Dana ('46), Gay ('54), Verrill ('68), 
and Ridley ('81), but unfortunately the descriptions furnish no suffic- 
ient basis for the identification of alcoholic material. The form 、、 hich 
Verrill ('68), with some reservations, refers to Lesson's Actinia nivea 
seems to be rather closely related and may be identical, though I should 
be inclined to doubt, without good evidence, the identity of a shallow 
water form with one living at a depth of 450 fathoms. It m doubtful, 
too, whether VerrilFs form is really Sagartia {Act,) niveau since Lesson 
expressly states that in this form "I'euveloppe est tres-lisse, tr^s-douce 
au toucher et seulment uiarqu<5e de quelques ondes ou plissures verti- 
cales," while Verrill ,s form has the "integument thin but firm," more 
nearly resembling S. lactea in this respect. It is on account of this un- 
certainty of definition that I have preferred to consider the Albatross 
form a new species. 

No. 954a. Station, 2764. Lat. 36^ 42' S. ; long. 56。 23' W. Depth, Hi fathomR. 
Three specimens. 

The three specimens differ somewliat in external api)earaiiee. One 
is quite small, while the other two were larger, measuring about 
0.5cm ill height and 0.6"" in diameter. One of tbe specimens was colored, 
the column being chocolate-brown in color, the tentacles much darker, 
but of about tbe same color. The other two specimens showed no 
traces of this coloration and may possibly be different species. The 
anatomical details given below were derived from the study of one of 
the colorless specimens. 

The base is adherent and not much larger than the column (PI. xxix, 
Fig. 77). This is somewhat wrinkled by contraction, but bears no warts 
or verrucae. Its wall is rather thin, soft, not parcliment-like. In one 
of the specimens an acontium protruded through the wall, but no eiu- 
clides were elsewhere visible. The sphincter (PL xxix, Fig. 78) Ls nar- 
row, but well developed. In its upper part it occupies the greater part 
of tbe thickness of the column wall and tapers off gradually below. In 
section the muscle-cavities in the upper part are eloii gated perpendicu- 
larly to tlie surface of the column, becoming gradually more circular 
towards the lower edge of the muscle. 

23. Sagartia Sancti Matthaei, sp. nov. 

Plate XXIX, Figs. 77 and 78. 


The tentacles are exposed to a greater or less extent in all the speci- 
meus; they are short, and pointed at the apex. They are strongly 
entacm^eous and their u umber is probably less than one hundred. 
Their longitudinal musculature is ectodermal and is fakly developed. 

The 】nesenterie8, as in the preceding species, are arranged upon a 
decamerous plan. There are ten pairs of perfect mesenteries consti- 
tuting the first and second cycles ; the third cycle is imperfect, but 
well developed, while the fourth cycle is considerably smaller. Here 
and there pairs of mesenteries of the fifth cycle can be seen, but this 
cycle is not com plete. No reproductive organs were present. The lou 
tudinal musculature is fairly well developed on the larger mesenteries, 
the mesogloeal processes increasing gradually in size towards the 
inner margin of the muscle and there abruptly diminishing. 

24 Sagartia paradoza, sp. no v. 

Plate XXIX, Figs. 79-81 ; Plate xxx, Fig. 84. 

No. 692. Station, 2766. Lat., 36。 47' S. ; lotag., 56^ 23' W. Depth, 10^ fathoms. 
Several specimens. 

In this form (PI. xxix, Fig. 79) the base is adherent. The column is 
longitudinally ridged with fine elevations, and does not bear any tuber- 
cles or vermcae, nor were any cinclides observable, acontia being 
emitted from the mouth, however, iu several specimens. Nearly all the 
specimens have the tentacles and disc perfectly iinretracted, and the 
stoiuatoddeum is more or less evagiuated in many. The specimens 
have an average height of about 0.8c™', and a diameter of about 0.9彻, 
Tlie sphincter muscle (PI. xxix, Fig. 80) is very well developed, notwith- 
standing the nonretractiou of the tentacles. It occupies the entire 
thickness of the inesogl(Ba, and is thickest about the middle, tapering 
oft above and below. The muscle cavities are very numerous, and are 
separated only by very thin trabeculie of mesogloea, so that the column 
wall ill the region of the sphincter has an opeuly reticulate appearance 
in longitudinal section. 

The tentacles occupy the margin and are very numerous, short and 
acuminate, and decidedly entacmaeous. Their longitudinal muscula- 
tuie and the radial musculature of the disk is ectodermal, the muscle 
processes being fairly well developed. The disc is smooth. The stoma- 
todseiim is longitudinally ridged, and has two siphonojj^lyphes, one of 
which, however, seems to be much deeper and more distinct than the 

The mesenteries present a rather peculiar arrangement in the two 
specimens of which an anatomical study was made (PL xxx, Fig. 84). 
They are arranged on an octamerous plan. If we consider for conven- 
ience in description eight pairs as constituting tlie first cycle, then the 
first three cycles (i, ii, iii) are all perfect, the mesenteries of the third 
cycle losing their connection however with the stomatodaeum about 

Digitized by 



half-way down. A fourth cycle of imperfect mesenteries is present, 
but it is not complete. There are two pairs of directives (D), one ot 
which, connected with the deeper siphonoglyphe, is 】imcli stronger than 
the other. The regularity of development of the mesenteries is some- 
what interrupted on either side of these smaller directives. Disre- 
garding the rudimeutary meseuteries of the fourth cycle, a pair of 
mesenteries (x) which are perfect succeed on each side these directives, 
and next there comes a pair {y) belonging to the second cycle, which 
consists of one perfect and one imperfect mesentery, the latter being 
nearest tlie directives. This arrangement occurred on both sides of 
the directives, and iu both tbe specimens examined, and accordingly 
is probably normal. 

.Acoutia are present, as above stated. All tho mesenteries ex(m0 
those of the fourth cycle and the directives are gonophoric. The lon- 
gitudinal musculature is well developed (PL xxix, Fig. 8), a marked 
pennon being present, the various muscle processes of which arise in- 
dependently from the inesogloea. 

No. 718. Station, 2799. Lat., 8" W N. ; long., 79 じ 09' W. Depth, 29i fathoms. One 

Tlie S, crispata descnoed by Verrill ('68) was dredged in from 4 to 
6 fathoms in Panama Bay, and occurred upon the shell of a large Miirex 
(Phyllonotus). The specimen which I identify witli it with some hesi- 
tation, was found in sliglitly deeper water in tbe same locality, and also 
occurred upon the shell of a good sized Prosobraiieli, apparently one 
of tlie Muricidoe. It is very much flattened iu contraction, the margin 
and tentacles being completely concealed. Tbe base measures about 
2.7cn' in diameter, and firmly dasps tlie surface of the shell, wliicli was 
inhabited by the living 動 Husk and not by a Pagurid. Tlie column is 
wrinkled and somewhat rougheued by iiiiiiiite elevations i>roduced by 
contraction, but does not seem to possess any vemicae. Acoutia are 
emitted through the colunni wall a short distance above the limbus, 
but no series of cinclidal tubercles could be made out. The column is 
marked by numerous, ii regularly wavy, longitudinal lines of a choco- 
late brown color, which are very distinct upon the white ground. 

Not wishing to destroy the single speciineu I can give no particulars 
regarding the internal structure. 

From the fact that the acontia are emitted a short distance above 
the limbus it is possible that this form is Adamsia. Its identifica- 
tion with 8. crispata is necessarily uncertain, owing to there being no 
opportunities for a thorough comparison of tbe two forms. The differ- 
ences between tbe coloration iu this form and Verrill's description of 8. 
crispata may be due to preservation. 

25. Sagartia crispata (Bradley) Verrill. 



Genus ADAMSIA Forbes. 

Sagaytinse with adherent base, the ectoderm of which secretes a 
membrane; column without warts or verruca?, but provided with one or 
two horizontal series of ciiiclidal tubercles a short distance above the 
limbus; margin tentaculate. 

No. 716. Station, 2793. Lat. 1° 03' N.; long. 80。 15' W. Depth, 741 fathoms. 
Twelve specimens. 

Every specimen is fully contracted and completely incloses a Gas- 
teropod shell, being wrapped around it iu such a manner as to conform 
itself more or less to the shape of the shell (PI. xxix, Figs. 82 and 83). On 
this account it is difficult to give any accurate measurement of the 
height of the Actinian, but this 】nay be averaged for the contracted 
specimens at about 1.5 to 2cm. , g j^j the diameter at from 1.25 to 1.5c™. 
The column is of a pale flesli color, but becoming thinner toward the 
limbus it has a darker shade, and is here longitudinally streaked with 
white lines, indicating the lines of insertion of tlie mesenteries on the 
column wall. Iii this thinner region, too, the internal organs shiue 
througli. The tentacles are of a salmon color, this tint depending, to a 
certain extent, and probably entirely, on tlie bright reddish orange 
pigment which occurs everywhere in the eudoderm. The coloration 
Avliich these preserved specimens present is entirely independent of any 
colors which may liave been present in the ectoderm, since this layer 
has entirely disappeared from tbe surface of the column. 

The base incloses the gasteropod shell, and, as it were, forms the 
openiug of the habitation of tlie mollusk. Its ectoderm secretes a 
very well-marked cliitinous layer, not only over tbe region iu contact 
with the shell, but also over that wliicli is free from it. 

The column wall is smooth tliroughout and lias a parchment-like con- 
sistency, the mesoglcBa being very fibrous in structure, though rather 
til ill. No cinclidal tubercles could be perceived. Tlie.spliincter muscle 
(PI. XXX, Fig. 85) is w ell developed, though not very broad. Toward its 
upper margin tlie muscle cavities are in section more or less circular 
ill outliue and distinctly separated from one another, but lower down 
tliey are more elongated and are separated by narrower partitions, cir- 
cular scattered cavities lying upon the outer surface. It is separated 
throughout from the endoderm by a thin layer of inesoglcBa. The circu- 
lar muscles of the endoderm are only slightly developed, the cells being 
arranged in an almost smooth layer and not supported on well-devel- 
oped processes of mesogloea. 

The margin is tentaculate, and the tentacles are arranged apparently 
in three cycles, though their exact arrangement it is difficult to ascer- 

26 Adamsia (?) involvens, ep. nov. 

Plate XXIX, Figs. 82 and 83; Plate xxx, Fig. 85. 


tain on account of tbe contraction of the specimens. They seem to be 
ninety-six in number, and to bo arranged in two cycles of twenty-four 
each, and one of forty-eight. Their longitudinal musculature is well 
developed and is entirely ectodermal, supported on strong mesogloeal 
processes. The mesogloea of the tentacles does not partake of the 
fibrous structure of that of the column wall, but is hyaline. 

Two siphonoglyphes are present, apparently, and two pairs of direct- 
ives. There are forty-eiglit mesenteries, only the six primary pairs 
being perfect. The secondary amf tertiary pairs bear the reproductive 
elements, those of tbe fourt h cycle being quite small and destitute of mes- 
enterial filaments. The longitudinal muscle processes are fairly well 
developed, but do not form a very distinct muscle pennon. Acontia 
are present; in some of the speciiiieiis tUey were emitted from the 
mouth, but in none did I fiud tlieiu protrudiug from the column wall. 

On account of any failure to discover cinclides it is of course doubt- 
ful if this form is correctly referred to tlie genus Adamsia. Tbe shape 
of the sphincter is decidedly 《liffm で iit from that of Adamsia parasitica, 
and A. polypus as described by Hertwig, but does not, however, differ 
so materially from that of Adamsia Sol of our eastern coast. My prin- 
cipal reasons for considering A. involvens a possible Adaimia is its hab- 
itat on gasteropod shells and the secretion of a strong chitiiious mem- 
brane by the ectoderm of the base, features which are, however, of com- 
paratively small value. 

Subfamily (Jhondraotinin も Haddoii. 

SagartidiB with thick column wall, usually with the upper portion 
(capitulum) different in character from tbe lower (scapus) and capable 
of being entirely invected ; the scapus pi'ovided with an external cuti- 
cle and usually nodulated or warty ; the sphincter strong- and imbedded 
in the mesogloea; only the six primary pairs of mesenteries perfect 
and at the same time nongonophoric ; acontia emitted by tbe mouth 
only, there being no cinelides. 

Genus ACTINAUGE, Verrill. 

Clioiidractiaiuab in which the capitulum is provided with lougitudiual 
ridges; scapus strongly tiiberciilate or nodulate, the tubercles near the 
junction of the scapus and capitulum being usually stronger than those 
lower •down; each teiitaelo with a bulbous tliickciiing on tbe outer 
surface at the base. 

This genus was established by Verrill ('83) to receive a form whicU 
he believed to be identical with the Aciinia nodosa of Fabriciiis. The 
definition given above contains the essential points of VerrilPs defini- 
tion, with the addition of a mention of the presence of a bulbous en- 
largement at the base of the tentacles, a feature to which H addon 
('89) has called atteiitiou, and made an important factor in the limita- 

Digitized by 


tion of the genus. In the definition given by Haddon the capitolai* 
ridges are limited to twelve, while Yerrill expressly states that tbey 
are as numerous as the tentacles. From an examinatiou of specimeBS 
of the type species I can state positively that there are forty-eight cap- 
itular ridges in it, one ridge corresponding to each of the twenty-lbur 
more or less distinct rows of tubercles, wLile a smaller ridge intervenes 
between each pair of these larger ones. Haddon likewise limits the 
bulbous enlargements to the bases of the three inner circles of tentacles, 
but ill the type species there is no such limitation in their distribution, 
all the tentacles possessing the enlargements. The numerical limitar 
tious of the ridges and bulbous eiilargemeuts must be regarded as of 
specific but not of generic value. 

Plate XXX, Figs. 86-89; Plate xxxi, Figs. 90-92; Plate xxxv, Fig. 121. 

Synonyms : Urtidna wotiosa, Verrill (1873) ; J ctinauge nodosa, Verrill (1883) ; Actinauge 

(sp.)?, Haddon (1889). 
No. 712. Station 2791. Lat. 38^ 08' S.; long. 75^ 53' W. Depth, 677 fathoms. 

Seven specimens. 

No. 734. Station 5839. Lat. 33^ 08' N.; long. 118" 40' W. Depth, 414 fathoms. 
One specimen. 

Nos. 733, 735. Station 2839. Lat. 33^ 08' N. ; long. 118^ 40' W. Depth, 414 fatlime. 

Six specimens, young. 
No. 724. Station 2818. Lat. 0〕 29' S. ; long. 89 〕 W 30" W. Depth, 392 fathoms. 

One specimen (much toi'ii). 

I have been able, by direct comparison, to identify the specimens 
marked No. 712 with specimens of A, Verrillii from the eastern coast of 
North America and shall give a detailed account of the structure of 
these specimens. The specimen No. 734 presents some differences from 
the typical A. Verrillii, and it is possible that it may belong to another 
species, but I did not investigate the structure of the single specimen, 
and will content myself with giving a description of its external pecul- 
iarities. Nos. 733 and 735 were obtained in the same dredging as No. 
734, and are probably young forms of the same species, and call for a 
brief description. Finally, No, 724 is referred to this species with some 
hesitation ; it is very much distorted and torn, so that it is impossible 
to examine it satisfactorily. It is possibly the tuberculosa variety which 
Verrill has described as a distinct species, but nothing can be said con- 
cerning it. • 

All the specimens of No. 712 are thoroughly contracted, the tentacles 
and capitulum being concealed (PL xxx, Fig. 89). The column is cylin- 
drical, and covered with well-marked, large tubercles, thickenings 
of the inesoglcBa, which are especially high iu the upper part of the 
coluniD, where they are somewhat square in outline, and arranged 
more or less definitely in horizontal and longitudinal rows, there being 
about twenty-four of the latter. Lower down upon the column the 

27. Actinauge Verrillii, no v. nom. 

VOL. xn,"i 

1893. J 



tubercles become much flatter, and toward the base they are repre- 
sented by slight transversely elongated, narrow elevations, the longi- 
tudinal arrangement being nearly lost. Tbo liinbus is smooth, the 
elevations fading out a short distance above it. 

The base is much smaller than the column and is deeply concave, a 
quantity of mud, which evidently served to anchor the animal, being 
inclosed in the concavity. 

The upper part of the column or capitulum does not possess any 
tubercles, these being limited to the scapus. The uppermost tuber- 
cles are usually more pronounced than those lower down, and form a 
more or less distinct coronal series (PI. xxx, Fig. 89 ear), consisting of 
twenty-four tubercles. From each coronal tubercle a ridge (c, r.), ex- 
tends across the capitulum toward the bases of tlie tentacles, and be- 
tween each pair of these coronal ridges a smaller ridge intervenes, so 
tbat the capitulum bears in all forty-eight ridges. Tliey are decidedly 
prominent, with thin walls, the cavities which tliey contain communi- 
cating with the endocoels. Before reaching tlie level of the bases of 
the outermost tentacles each ridge somewhat suddenly increases in 
height, and more suddenly diminishes, giving? rise to a pouch-like 
structure. The ridges terminate at the bases of the tentacles of the 
four inner cycles, tlie tentacles of the outer cycle being situated upon 
the sides of the intermediate smaller ridges, in tlie manner indicated in 
tbe scheme given on PI. xxxv, Fig. 121. 

Wlien the cuticle is preserved tbe colunni lias a dark-browii color, 
but the tubercles are white for the most part, owing to tlie cuticle hav- 
ing been rubbed off. The capitalum in tbe alcoholic si)eciineus is 
colorless 5 the disc and tentacles, however, are orange or salmon 
colored, while the Btomatodieura is brown. 

The mesoglcBa of the column wall is thick and delicately fibrous in 
structure, with n. few cells scattered through it. The ectoderm, where 
present, is covered by the thick cuticle, to which particles of foreign 
matter adhere. Tbe tubercles are solid elevations of the mesoglcea. 
The sphincter muscle (PI. xxx, Fig. 86) is fairly strong, but varies some- 
what, both in its thickness and width, in different specimens, the dif- 
ferences not being due to age, as in some cases I have found tbe muscle 
mnch weaker in a large specimen than in smaller ones. It occupies the 
entire capitular region, and extends a varying distance below the cor- 
•onal tubercles. Throughout its entire width it is widely separate from 
the endodermal surface of the column and lies in the scapus very close 
to the ectoderm. In transverse section (PI. xxx, Fig. 88) it is seen to 
consist of more or less circular cavities, traversed by irregular par- 
titions of mesoglcea, though in some cases the cavities are more numer- 
ous and smaller, and almost destitute of partitions. Toward the lower 
edge of the muscle the cavities are in one or two series, but they be- 
come more numerous above, but there is no well marked, sudden thick- 
euing of the muscle in its upper part. A curious arrangement is found 


in the upper part, in some forms at least; the muscle fibres and the 
cavities instead of being cut across by a transverse section through 
the muscle, give the appearance of being cut parallel to their course 
(PL XXX, Fig. 87) and the section lias the appearance of a horizontal or 
transverse section through the upper part of the column wall. Further- 
more, the cavities, brauchiug and anastomosing witli each other, pass 
toward the ectodermal surface of tlio mesogloea, and apparently in 
some cases come into contact with the ectoderm. This arrangement, 
as I have said, is not so distinct iu some specimens as in others, but is 
more or less marked in all my preparations. 

The tentacles are ninety- six iu number and are arranged iu five 
cycles. They are rather short, but slender and pointed. At the outer 
surface of the base of each there is a bulbous swelling (PI. xxx, Fig. 89), 
formed principally of thickened mesogloea (PL xxxi, Fig. 91). The ecto- 
dermal musculature, both of the tentacles and of the disc, is rather 
weak, the uiesogloeal process for its support being only slightly de- 

The stomatodaeum is long, extending, iu tlie contracted specimens, 
almost to the base. It lias two sipboiioglyphes, which are well devel- 
oped though not particularly deep. 

There are tweuty-four pairs of mesenteries arranged iu four cycles. 
Only the six mesenterial pairs of the first cycle are perfect. The 
mesenteries of the second cycle, though imperfect, resemble those of 
tlie first cycle in being nongonophoric, the reproductive organs being 
borne altogether by the mesenteries of the third (PI. xxxi, Fig. 90) and 
fourth cycles. In the region of the mesentery occupied by the repro- 
ductive elements iu female individuals the mesogloea is greatly enlarged 
(PL XXXI, Fig. 90), the ova (ov) being imbedded in the enlargement. 
This does not occur in tbe mesenteries of male individuals from the 
Atlantic coast of North America; all the Albatross specimens I ex- 
amined for this point proved to be females. The longitudinal muscles of 
the mesenteries are not particularly well developed (PI. xxxi, Fig. 92), 
and thereisnocircumscribed pennon. The low mesoglceal processes tend 
somewhat to be arranged in bunches of a few arising from a common 
basis. At the bases of the mesenteries, i. e" at their attachment to the 
column, there is a well-marked pinnate parietal muscle. The acontia 
are not abuudaut. 

No. 734, as stated above, differs in some respects from No. 712. Its 
base is not deeply concave as it is in No. 712, nor does it seem to 
have inclosed mud or sand for an anchor, but appears to have been ad- 
herent. The tubercles of the column are somewhat more distinct and 
rounded than in No. 712, and are all covered by cuticle. The sphincter 
has essentially the same structure as No. 712, but I did not dissect the 
specimen sufficiently to determine if the likeness extended to all the 
parts. I think, however, that there is little reason for disbelieving in 
the specific identity of the specimen with No. 712. 

VOL. xy 



The specimens Nos. 733 and 735, obtained in the same dredging as 
NcK 734, are both small, and white or pale brown in color, the cuticle 
being only very slightly developed. The base is only slightly concave 
and seems to have been adherent. The upper part of the column is 
marked by twenty-four longitudinal ridges, which show more or less 
distinct traces of transverse grooves, dividing each ridge more or less 
perfectly into a series of tubercles. I see no reason for supposing that 
these are other than young individuals of the same species as No. 734. 

In changing the name of this species I have followed the suggestion 
made by Prof. Haddon ('89), and have named it after the distinguished 
naturalist who first described it. Verrill ideuiitied it with the Actinia 
nodosa of Fabricius, but the more recent observations of Haddou ('89) 
and DauielHseu ^90) show that the two forms are quite distinct, and 
the former has assigned Fabricius' form to the genus Chondractinia pro- 
posed for it by LUtken ('60). This being the case, it seems advisable, 
for the avoidance of the confusion which might ensue from two so 
closely related forms possessing the same specific Dame, to change the 
name of VerrilPs species. 

Synonym. 一 Actinauge nodosa, var. coronatOf Verrill (1883). 

No. 713. Station 2791. Lat. 38^ 08' S. ; Ion も lo^ 53' W. Depth, 677 fathoms. 
Seven specimens. 

It is customary, when a form originally described as a variety is ad- 
vanced to the dignity of a species, to employ the varietal designation as 
the specific uame. I have thought it well in the present case to depart 
from this precedent, since the specific term coronata has already been 
applied to a form belonging to the genus Chitonactis, which is nearly 
related to Actinauge. 

The specimens of Actinauge fastigata, obtained by the Albatross ^TOln 
the same locality as most of the specimeus of A, VerrUlii, are in all re- 
spects similar to those described by Verrill ('83) from deep water off 
the St. George's Banks. 

The specimens measure from 3.5 to 4.4«"' iu height, with a diameter 
at the upper part of the column of from 2 to 2.5ゅ、 

The base is somewhat smaller than the column and, apparently, is 
adherent; one of the specimens clasps the tube of a Hyalincecia. In 
none of the specimens is it deeply concave, inclosing mud or sand, as 
is the case with A. VerrillH, The limbus is smooth, and iu nearly all 
the specimeus is destitute of cuticle and is rather thin, allowing the in- 
sertions of the mesenteries to show through. 

The column is cylindrical (PI. xxxi, Fig. 93), gradually increasing in 
diameter tawards the upper part, the capitulum being, however, com- 
pletely infolded in all the specimens. The lower part of the column is 
covered with low and small warts, arranged, more or less distinctly, in 

28. Actinauge fastigata, uom. nov. 

Plate XXXI, Figs. 93-97. 


rows, and giving the column almost a granular appearance in some 
specimens. They become smaller as they approach the limbus, and 
fade out a short distance above it. A dark brow ひ cuticle covers this 
portion of the column. Just below the capitulum, and forming there- 
fore the summit of the contracted column, are two circles of very 
prominent tubercles, tipped with blunt cliitinous points. There are 
twenty-four such tubercles, arranged in two horizontal rows of twelve- 
each, so that there may be said to be twelve longitudinal rows of these 
large tubercles, each row consisting of two tubercles. Between each 
pair of longitudinal rows there is usually to be seen a row of small 
tubercles, so that there are in all twenty-four longitudinal rows of 
tubercles, twelve of them being very large and prominent, and twelve 
small and almost bidden by the larger ones. The capitulum is essen- 
tially the same as that of A. Verrilliiy possessing forty-eight longitudi- 
nal ridges which run to the bases of the tentacles. It is destitute of 
cuticle and tubercles. The sphincter (PI. xxxi, Fig. 94) resembles that 
; of A. Verrillii closely. It is tolerably wide, but not thick, being only 
slightly thicker in its upper part than it is lower down. In section it 
appears as a number of more or less circular cavities, traversed by del- 
icate partitions, which support the muscle cells. In the lower part 
(PI. XXXI, Fig. 95) there is only one sucli cavity to the thickness of the 
muscle, but above (PI. xxxi, Fig. 96) there may be three or four, or even 
more, since the cavities tend to become smaller in the upper part. 
Tliroughout its whole widtU the muscle is separated by a broad band of 
mesogloea from the endodermal surface of the column, lying nearly mid- 
way between the two surfaces. 

The tentacles are ninety-six in number, as calculated from the 
Dumber counted in a sextant. They are decidedly eutacmaeons, and 
are arranged apparently in four cycles, it being difficult to distiDgaish 
those of the first two cycles by their position. Each tentacle possesses 
at its base a. bulbous enlargemeut similar to that described for A, 
Verrillii. The longitudinal muscles of the tentacles are weak. The 
tentacles, disc, and stoinatodseum seem to have been of a salmon or 

The stomatodaeuni is provided, with two rattier shallow siphonogly- 

The mesenteries are arranged in three cycles, there being only 
twenty-four pairs in the specimen examined. Probably, however, a 
fourth cycle is present in larger specimens, since the number of 
tentacles would lead one to expect forty-eight pairs of mesenteries. 
The mesenteries of the first cycle are perfect and uongoiiophoric, those of 
the other two cycles being imperfect and at the same time gonoplioric. 
The loDgitudinal musculature is well developed (Pi. xxxi, Pig, 97), there 
being a strong muscle pennon situated near the outer edge of tbe 
mesentery and having a somewliat abrupt inner edge, beyond which, 
however, are a number of much lower muscle processes gradually di- 



minishing iu size and finally disappearing a little internal to the mid- 
longitadinal line of the mesentery. 

As stated above, Verrill originally described this form as a variety 
of A. Verrilliiy stating that intermediate states between it and the normal 
form are not rare. The Albatross specimens do not show any such 
intermediate gradations, though botli the presumed variety and the 
type species were obtained from the same locality. Leaving out of 
consideration the possibility of an approximation of the arrangement of 
tlie tubercles in the two forms, there are yet other characters wbich, it 
seems to*me, are of sufficient imi)ortance to necessitate the separation of 
the two forms as distiuct species. These may be briefly summed up as 
follows : (1) The proportion of the diameter to the height of the column 
in A. fastigata is considerably less than in A, Verrilliiy the latter having 
consequently a much more robust form than the former; (2) the base iu 
A. fastigata is adherent, while in A. Verrillii it is deeply concave and 
incloses a mass of mud or sand which serves as an anchor; (3) the 
relations of the nongonoplioric and gonophoric mesenteries differs iu 
the two forms; (4) the longitudinal musculature of A, Verrillii is weak, 
whereas in A, fastigata it is strong and formK a well-developed pennon. 

Gliondiactiniiise in which the capitulum is provided with longitudinal 
ridges ; the scapus, especially iu its upper portion, with strong pointed 
tubercles not arranged iu any definite order, or else with a single circle 
of coronal tubercles; the cuticle strongly developed upon the tuber- 
cles; tentacles without auy bulbous enlargement at the base. 

I suggest this geuus for two forms already described by Hertwig 
('82, ,88) as Fhellia pectinata and Phellia spinifera. There can be no 
doubt that it is advisable to remove them from the geuus Phellia, the 
typical members of which liave a smooth capitulum. If the detinitions 
which Haddou ('89) has proposed for the various genera of Chou- 
dractinidae be accepted, Hertwig^s Phellia spinifera finds no place 
among them. It comes close to Chitonactls, but differs in possessing 
ridges upon the capitulum. It is to be noticed that Haddon has as- 
signed the form described by Hertwig ('82) as Phellia pectinata to tlie 
genus Hormathia of Gosse. If this be correct, Fhellia spinifera must 
be referred to the same geuus whose definition will require to be 
amended so as to include forms possessing tubercles scattered irregu- 
larly over the scapus. However, if the figure given by Gosse ('60) of 
his Hormathia margaritce be correct, its capitulum is smooth and it 
would perhaps be as well, especially when we consider how little is 
definitely known regarding the type species of the genus, to reserve 
Hormathia for those forms in which the capitulum is smooth and which 
possess only a coronal row of tubercles, associating the Fhellia pectinata 
of Hertwig and the Hormathia andersoni of Haddon ('88), which 

Genus CHITONANTHUS, gen. dov. 


possess only coronal tubercles but have a ridged capitulum with Hert 
wig's Phellia spinifera in the new genus Chitonanthus. It is of coarse 
a question as to whether the presence or absence of capitular ridges is 
worthy th も importance which this arrangement gives it; but it must be 
recognized that the classification of the Chondractininw is at present 
more a question of convenience in identification than of phy logenetic re- 
lationship, and that what may be trivial characters have been raised to 
the elevation of generic distinctions. Thus, to judge from Haddon's 
definitions of the genera, the principal feature which distingnisbes 
Chondrmtinia from Chitonactis is that the tubercles in the latter are 
pointed, while they are mostly low and nodule-like in the former. (See 
appendix p. 209.) 

SynoDym : Phellia pecHnaia Hertwig (1882) ; Phellia spinifera Hertwig (1888). 
No. 703. Station 2780. Lat. 53。 01' S. ; long. 73° 42' 30" W. Depth, 369 fathoms. 
Three specimens. 

The three specimens which represent this sjiecies have a very different 
appearance from one another. One (PI. xxxii, Fig. 98), which may be 
considered the most typical, is seated upon a detached valve of a Larael- 
librau (; h shell by a broad, flat disk. Its column was much contracted 
and thrown, to a certain extent, into folds. It measured 2:1«"» in height 
and I.IK m in diameter, and was covered with irregularly scattered tuber- 
cles whicli were low and flat near the base, but sharp and prominent 
above, where they become more numerous. The upper tubercles owe 
their sharpuess to a strong development of cuticle over them, and it is 
possible that in the lower ones this cuticular point has been lost 
Though scattered irregularly over the column for the most part, yet 
they show a tendency to arrange themselves above in twelve longitu- 
diual rows. 

The second specimen, the one which I chose for detailed study, is 
larger than the first, measuring 3.5彻 in height aud 3*^" in breadth. Its 
base is broad and flat, like that of the first specimen, but bad been de- 
tached from its support, only particles )f a shelly nature being attached 
to it. The column is almost smooth and white in color, the brown cuti- 
cle, which covered the first specimen, having disappeared, except in the 
immediate neighborhood of thelimbus. The general smoothness of the 
column is, however, relieved by a few nodule-like elevations (PI. xxxii, 
Fig. 99), and some rarer, more prominent 】iodules tipped with brown 
cuticle. Toward the summit, however, one finds twelve strong ridges, 
each more or less broken into rows of tubercles and terminating above 
in a strong tubercle tipped with a prominent thickening of cuticle. 

The third specimen measured 2<"" in height and 2.5^™ in breadth, and 
was seated upon the valve of a Lamellibranch shell. Like the second 
specimen it was white in color, only a few isolated patches of cuticle 

29. Chitonanthus peotinatus (Hertwig). 
Plate XXXII, Figs. 98-102. 


1893. J 



persisting. It differs from both the others, however, in being utterly- 
devoid of tubercles, the only indication of any sucli structures being the 
occurrence of about twelve ridges at the upper part of the column, 
which end abruptly at the junction of the capitulum and scapus, but 
are not tipped with a ciiticiilar thickening. 

The external appearance of these three forms is so dissimilar that 
one might suppose them to be distinct species. Their occurrence in 
the same locality, the similarity of their support, in each case a Lamel- 
libranch shell, and the gradations which they show led me to believe 
that they were identical. I made a detailed study of only one, the 
second, and consequently can not speak as to the identity throughout 
of the internal structure, but so far as this could be examined by slit- 
ting the specimens longitiidiDally there was perfect similarity and I 
have little doubt but that all three ought to be assigned to the same 

The infolded capitulum in all the specimens possesses twelve longi- 
tudinal ridges and, as in Hertwig's Phellia pectinata, the ridges towards 
their upper termination are divided by a longitudinal furrow which 
maybe extensive enough to give the appearance of twenty-four ridges. 
In the first and second (PI. xxxii, Fig. 99) specimens a few tipped tuber- 
cles are found on the infolded portion of 'the column, resting in the lower 
portion of the ridges, and each is more or less distinctly cleft into two 
parts. .The strong sphincter (PI. xxxii, Fig. 100) has the general appear- 
ance figured by Hertwig for P. pectinata. In its lower part it is thin and 
composed of cavities which are circular iu section, but in its upper part 
(PI. XXXII, Fig. 101) it thickens somewhat and the cavities are elongated 
in a direction perpendicular to the surface of the mesogloea, some 
scattered round cavities occurring upon the outer surface of tb^ muscle. 
I did not find in the mesogloea of the column wall any of the concre- 
ments which Hertwig describes in P. pectinata. These seem to have 
been absent in his P. spini/era and are probably accidental foreign in- 

The tentacles (PL xxxii, Fig. 99, り are rather short and slender and do 
not api)ear to have a bulbous enlargement at the base. They are ar- 
ranged in about three cycles and appear to number forty-eight. The 
first two cycles correspond to the ridges of the capitulum, regarding 
each of these as really representing two, while the third cycle tentacles 
alternate with the ridges. The lougitudinal muscles of the tentacles 
are fairly well developed and are not imbedded in mesogloea. In color 
the tentacles seem to have resembled the disc, which was of a purplish 
brown color. Its radiating muscles present tbe peculiarity already 
described by Hertwig iu P. spini/era. 

The stomatodaeum is long, reaching to below the middle of the inter- 
nal cavity (PL xxxii, Fig. 99, "•), and is of the same purplish* brown 
color which marked the tentacles and disc. The broad but shallow 


siphonoglyphes are, however, not pigmented, and consequently are 
very noticeable when the animal is opened longitudinally. 

There are four cycles of mesenteries, of which the primary cycle is 
alone perfect, and at the same time sterile; the fourth cycle mesente- 
ries are small and are not gonophoric, the reproductive elements de- 
veloping only in the mesenteries of the second and third cycles. The 
longitudinal musculature is well developed (PI. xxxii, Fig. 102), bat 
can hardly be termed "very strong." The peimon is not wide, the muscle 
processes arising in bunches from one to three stout elevations of the 
mesogloea; it is much more marked in the upper portions of the mesen- 
teries than it is lower down, where it becomes lower and at the same 
time broader. I did not observe any extensive folding of the trans- 
verse muscles, nor could I find in sections any parieto basilar muscle. 
Acontia are present, lying iu bunches in the lower portion of the inter- 
nal cavity. 

I identify tbis form with Hertwig's Phellia spinifera^ with which it 
agrees closely. I have, however, accepted the possibility which Hert- 
wig suggests, that his P. spinifera may be a variety of liis P. peetiTiataj 
described in his first report ('82). The dissimilarity in the arrange- 
ment of the tubercles in the two forms is to a certain extent, as he re- 
marks, bridged over by the specimen obtained from station 320, and 
the second and third Albatross specimens help to bring the two forms 
into closer connection. If the difference in tbe nature of the disc mus- 
culature ill the two forms holds throughout, it maybe necessary to con- 
sider them distinct, but since iu all other particulars they shade into 
each other so closely, I think it better to consider them for the present 

Chondractininae iu which the body is elongated in the transverse 
axis, the base inclosing a cylindrical body, such as an Alcyonarian 
stem; column with thick wall, but not covered by a well marked 
cuticle; capituluui smooth, separated from the smooth scapus by a 
well marked circular swelling. 

In bis report on the Actiniaria, obtained by the Challenger, Hertwig 
('82) established a family Amphiaiitbidae for two genera Amphianthm 
and StephanactiSy both of wbich were characterized by the body being 
transversely elongated, the base clasping and inclosing the stem of a 
Gorgoiiia. From his observations on Stephanactis ttiherculata he found 
that in the arrangement of the mesenteries, and in the presence of a 
sphincter muscle imbedded in the mesogloea, there was a great similar- 
ity to a Sagartid, but he failed to discover acontia, although cinelidal 
openings pierced the column wall. Previously to Hertwig's discovery 
of these forms, von Koch ('78) had described au Actiuian, adherent 
to and embracing by its base tbe stem of Isis elongataj and in this he 
fancied he had found a clue to the ancestry of the Antipatharia. This 

Genus STEPHANACTIS, Hertwig. 



form, which von Koch named Oephyra dokrnii, Haddon ('89) has iu- 
f vesijgated, and finds that " it belongs to the series of typical Sagar- 
tians." Dauielssen ('90) again has described a form Korenia margar- 
(tueea, probably more correctly assignable to Hert wig's genus Amphi- 
anthusj concerning which he states that " the gaatral filament^ are richly 
beset with nematocysts," a remark which suggests the presence of 
acontia. He, however, finds that there are twenty-four perfect mesen- 
teries, though acknowledging a possibility of error i 1 1 this determinatiou. 
Mention must be made also of VerrilFs Actinauge nexilis ('83), a super- 
ficial examination of which leads one to the conclusion that it is a 
Chondractinian, though I have not been able as yet to detect the occur- 
rence of acontia, the single specimen in my possession not being 
aatis&ctorily preserved, and consequently not suitable for accurate 
observation. A study of sections, which, uufortuuately, I have not yet 
been able to make, may reveal these structures. Coueerning this form I 
believe, too, that it is identical with Stephanactis abyssicola first described 
by Moseley ('77). It is undoubtedly a Stephanactis, and the superficial 
resemblance to Moseley ,8 form is so close that, relying ou the external 
characters, which are all in reality that we have to base a judgment 
I npon, one would have little hesitation in prououiicing in favor of the 
■ specific identity of the two foniis. Finally, Chitonactis marioni Haddon 
1 ^'89) resembles Stephanactis in the elongation of the transverse (!) 
i axis, and the clasping nature of the base, and is, fide Haddon, a Sa- 
[ gartian belonging to the subfamily Chondractininje. 
I In view of this evidence, which it must be acknowledged is by no 
I means conclusive, I think it is advisable to abolish the family Am- 
\ phianthidsB and include Stephanactis and Amphianthus under the sub- 
family Choudractiniase. Furthermore, it seems not improbable that it 
[ may be necessary to disregard the clasping habit, and the consequent 
elongation of the body to the transverse axis as generic characters, 
since, as in the case of Chitonactis marioni these features may be as- 
1 sumed by species belonging to genera not characterized by them. 

Independently, however, of these features depending on the habitat, 
the genus Stephanactis is sufficiently well marked, out from other Choii- 
dractiuidaB to warrant its retention. 

No. 720. station 2807. h^^. 0^ 24' S; ^Qi^g. 89° p6' W. pepth, 812 fathoms. One 

The single specijneflL I was unwilling tQ mutilate any more than was 
absolutely necessary, and consequently am unable to give an p^ccurate 
description of its structure; nor can I eyen determine from it lyjietbw 
pr not acontia present. 

30. Stephanaotiq hyalonematiB, sp. nqv. 

Plate XXXII, Fig. 163. 



The base clasped firmly a small bunch of Hyalonema fibres, the mar- 
gins of thebase coming into close contact, but still being separable by the 
use of a little force. The animal (PI. xxxii, Fig. 103) is much elongated 
in the direction of the axis of the bunch of fibres to which it is attached, 
and is low. The column wall is remarkable on account of its brittleness. 
It is hard and brittle, like parchment, and is much wrinkled by con- 
traction. In sections through a small piece of the wall no ectoderm or 
cuticle could be observed, but the niesogloea was found to have been 
altered into a chitin-like substance, not taking • the stain ( borax- 
carmine) except on the outer and inner surfaces to a slight extent. A 
distinct thickened ring surrounds the upper part of the column sepa- 
rating the scapus from the capitiilum. The latter has a slightly 
irregular surface, but is not tubercii】ate, and differs from the scapns in 
lacking the chitin-like induration of the mesogloea. I could discover 
no trace of cinclides. A strong sphincter imbedded in the mesogloea is 
present, but I can give no account of its shape in transverse sections. 

The tentacles are completely concealed, and iny preparations do not 
throw any light upon the number or arrangement of the mesenteries. 

ActininaB with numerous perfect mesenteries, and with a strong, cir- 
cumscribed endodermal sphincter. Column wall fiequently provided 
with tubercles, verrucae, etc.; margin frequently with complicated 
acrorhagi. No acontia. 

Hertwig (,83) established this genus for a form which differed from 
all forms which had previously been assigned to the family Bunodidae 
by lacking the tubercles or verrucae which had been considered char- 
acteristic of the family. The internal arrangements of the Challenger 
specimen showed it, however, to be closley related to the verrucose 
forms, with which Hertwig very properly associated it, substituting 
for previous definitions of the family, which he named Tealidae, a more 
accurate one founded upon an. anatomical basis. 

To the geuus Leiotealia are to be referred Biinodidfe without verrucae 
or acrorliagi, and perhaps to this may be added the pinnate arrange- 
ment of the muscle processes constituting* the sphincter. 

No. 702. station 2779. Lat. 53。 06' S. ; long. 70° 40' 30〃 W. Depth, 77i fathoms. 
t)ne specimen. 

The base is firmly adherent to a large annelid tube. The celumn is 
contracted to a somewhat conical shape, and measures 2cm height 
with a diameter midway between the base and margin of 2.3*^". It is 

Family BUNODID^. 

Genus LEIOTEALIA, Hertwig. 

31. Leiotealia badia, sp. nov. 

Plate XXXII, Fig. 104; Plate xxxiii, Fig. 106. 



1893. J 



. wrinkled transversely by contraction, aud also is roughened by im- 
meroas small elevations, which, however, do not represent tubercles or 
verrucae. The specimen is one of the few of the collection which have 
retained to a certain extent their coloration, the ectoderm of the 
column wall being of a dark brick-red color. A noticeable feature is 
the readiness with which the thick ectoderm separates from the meso- 
gleea in large pieces; an explanation of this is found in the peculiar 
structure of the lower layer of the ectoderm. The outer layer of the 
ectoderm contains a few 隱 all iiematocysts and a large number of 
gland cells, some of which staiu very deeply with carmine, while others 
hardly stain at all. Below these there is a granular layer which stains 
rather deeply, and next to the mesogloea, occupying the region, where, 
in the tentacles, for instance, the nerve layer is found, is a broad, 
slightly stained, fibrillar layer, in which are numerous delicate spindle- 

f shaped cells. It seems probable that this layer is more or less nervous 

1 in its nature, but whether it is to be regarded as entirely nervous can 
not be determined. It is in this layer that the rupture takes place, 

r when portions of the ectoderm detach themselves, the structure of the 

i layer being delicate aud maceration of it easily brought about. 

I The region of the column immediately above (or internal to) the mar- 
gin is mucli depressed, appearing to represent a distinct fosse, and at 
the bottom of the depression there is present a strong circumscribed 
endodermal sphincter ( PI. xxxii , Fig. 104). In section it resembles closely 
that which I have described ('80) for Discosoma anemone, consisting of 
a central axis from which processes arise, producing a more or less 
pinuate appearance. 

There being only a single specimen of the form, I cut out only a 
small portion oi it for the examination of the sphincter, a piece of the 
excised portion being cut transversely for a study of the musculature 
of the mesenteries. I can not accordingly give any facts as to the ten- 
tacles, disc, or stomatodaBum, or even regarding the arrangement of 
the mesenteries. A few tentacles were cut in making sections of the 
sphincter, and it was evident from these tliat their lougitudiiial 
muscles were very weak. 

. The small portion which was sectioned for the purpose of ascertain- 
ing the nature of the musculature of the mesenteries contained repre- 
sentatives of three cycles of mesenteries. Two of these bore repro* 
ductive organs, while the third was sterile. Approximately tbe ex- 
cised portion represented one-twelfth of the circumference, and it may 
be computed that there are at least tweWe sterile (and perfect?) pairs 
of mesenteries and twenty-four pairs that are gonophoric (and imperfect). 
The mass of the mesenterial filameuts is very great, but no acontia 
could be recognized. The musculature of the mesenteries is fairly strong, 
gradually increasing in thickness from near the parietal edge to about 
the middle of the mesentery, Avhere it abruptly diminishes (PI. xxiii, Fig. 
106); tlie parieto-b^silars (jpbm) form distinct folds upon the surfaces 



of the mesenteries, and iiumerous cavities are inclosed between the 
mesogloea of the fold and that of the mesentery proper, as in Aetinostola 

Owiug to the lack of more complete data with regard to this form, I 
at first hesitated to classify it. It seems, however, to belong to Hert- 
wig's genus Leiotealiaj though without some knowledge as to the ar- 
rangement of the tentacles, the correctness of this reference must 
remain uncertain; tbe probability seems to lie in favor of a cyclical 
arrangement of the tentacles. The form recalls somewhat that described 
by Verrill ('67) for Kagosima Bay, Japan, as Urticina coecinea^ but 
the "few, slightly prominent, inconspicuous verrucae^ could not be 


Hexactiniae with simple conical tentacles at some distance from the 
apparent margin; between them and the margin arc low tentacular or 
foliose structures (fronds). Sphincter endodermal, more or less circum- 
scribed, lying in tbe interval between the teutacles and the frondose 
or tentacular structures. From two to several cycles of mesenteries 

I have elsewhere ('89«) discussed the question as to whether tbis 
family should be referred to the suborder Stichoclactyliuae, as ADdres 
('83) has done, or placed in the suborder Actinia ae, and have decided 
in favor of the latter position. Upon this view the froHds are to be 
regarded as differentiated acrorhagi. 

Genus OULACTIS, M.-Edw. 

PhyllactidjB in which the column is provided with longitudinal rows 
of verrucie in its upper part; the fronds are foliose. Sphincter muscle 
more or less circumscribed. 

32. Oulactis califomica, sp. no v. 
Plate xxxii, Fig. 105; Plate xxxiii, Figs. 107-108. 
No. 741. Pichilingue Bay, Gulf of California. Two specimens. 

The base is adherent and rather thin, allowing the insertions of the 
mesenteries to be seen through it. The column (PI. xxxii, Fig. 105) is 
cylindrical, and in the alcoholic specimens shows no trace of color. The 
two specimens measure, respectively, 3。*" and 3.5cm \^ height, with a 
diameter near the upper part of the column of 2*^°* and near the base 
of 1.5cm. Toward tbe upper part of the column are verrucae arranged 
ill forty-eight longitudinal bows, each row being composed of from eight 
to ten verruccT. Tbe upper portion bearing the fronds is not concealed. 
The fi'onds occupy the margin and extend inwards to the bases of the 
tentacles, which surround tbe mouth ; they are foliose, apparently becom- 
ing thickly so toward tlieir external extremity, and appear to be forty- 
eight m aumber, cQrre^poftdmg to tbe xom Qi yerruce^, but owing to 

Digitized by 

VOL. XVI, 1 
1893. J 



their close approximation in the preserved specimens their exact num- 
ber could not be accurately determined. On the eudodermal surface of 
the region which intervenes between the fronds and the tentacles is the 
sphincter, whose form may be better understood from the figfure (PI. 
xxxni, Fig. 108) than from a verbal description. It will be seen that 
it approaches the circumscribed type, but still has a considerable attach- 
ment to the column wall. It may, however, be fairly termed circum- 

The tentacles are simple and few in number. They appear to be 
arranged in two cycles, there being six iu each cycle, but it is diffi- 
cult to make them out satisfactorily in the preserved specimens. 

The stomatodaeura is provided with longitudinal ridges supported on 
elevations of the mesogloea. The siphonoglyphes are deep, with smooth 
walls, aud with the inesogloea much thickened. There are twenty-four 
pairs of mesenteries, twelve of them being perfect. 'The longitudinal 
muscles form a broad, well-defined muscle pennon (PI. xxxiii, Fig. 107), 
and a well-developed parieto-basilar muscle is present. No reproduc- 
tive elements could be discovered. 

This form may have some relationship to the form described by Ver- 
rill ('68) . as LophaetiH ornata^ as in that form the. fronds are more 
foliose near their outer ends than toward the bases of the tentacles. 
They seem, however, to be more numerous, though, as stated above, it 
was difficult to decide upon the exact number, owing to their confusion 
with one another in the contracted Reserved specimen ; perhaps 
twenty-four would be more correct, each showing indications of a 
division into two toward the outer extremity and so giving the ap- 
pearance of forty-eight. It seems probable that it is unnecessary to 
separate Verrill's genus Lophactis from Oulactis, though very decided 
differences exist between the present form and his L. ornata^ with 
which one might be inclined to identify it. 

Phyllactidae with the fronds represented by bunches of simple 
slightly branched, short, tentacle-like structures. Sphincter aggre- 
gated or circiimscriDed. Column with verrucre. 

Among the actiniae which I described from the Bermuda Islands 
('89a) was one which I referred to the genus Oulactis as O. fascicu- 
lata. I propose here to unite this form, which differs markedly from 
Oulactis in the structures of its fronds, with a form in the Albatross 
collection, in the above new genus. An objection to this may be found 
in the very different nature of the sphincters in tbe two species, that of 
the one being almost diffuse, while the other is typically eircuin scribed. 
The structure of the fronds bas been a generic character hitherto for 
tlie PliyllactidiB, and it is convenient for the present to retain it as 
such. When the anatomy of a larger number of species is known, it 
cau be determioed whether a classification upon this basis can be re- 

Genus CRAOACTIS, gen. nov. 


33. Cradactis digitata, sp. nov. 

Plate XXXIII, Figs. 109-112. 

No. ^a. Station 2766. Lat. 36^ 4 で S. ; long. 56。 23' W. Depth, 10| fathoms. 
Three specimens, two of wnich, however, are small. 

The three specimens of this species were contained in the same 
bottle which held the forms described above as Sagartia paradoxa. 

The base of the single adult specimen was injured, so that it is impos- 
sible to say whether or not it was adherent originally. The column is 
cylindrical, and measures in the adult specimen 2«'" in lieight and LS^*" 
in diameter. The base is somewhat smaller than the column. Numer- 
ous, somewhat scattered, verrucae occur on the column wall, being inucb 
more distinct near the apparent margin than lower down. The fronds 
consist of buDclies of short, blunt, tentacle-like processes (PI. xxxiii, 
Fig. 110), each of which divides, near its extremity, into two short arms. 
The endoderm of the fronds is colored with brown pigment. The 
sphincter (PI. xxxiii, Fig. Ill) is very strong and is circumscribed, 
resembling closely that form of sphincter Avliich is characteristic of the 

The tentacles are short and stout, and each has apparently a pore at 
the tip (PL XXXIII, Fig. 109《). They seem to be arranged in about two 
cycles, and are not numerous, probably not exceeding forty-eight. 
Their endoderm contains brownish pigment similar to . that of the 

The stomatodaBuin in all the specimens is considerably evaginated. It 
possesses two well developed and deep siplionoglyphes, whose mesoglcea 
is decidedly thickened and smooth, that of the rest of the stomatodaeum 
being raised into longitudinal ridges. 

There are twenty-four pairs of mesenteries, twelve of which are per- 
fect. The longitudinal muscles (PI. xxxiii, Fig. 1 12) are well developed, 
forming a broad pennon, siniilar to that of Oulactis californica; the 
parieto-basilar muscle is also well developed, forming a fold upon tbe 
surface of the mesenteries. No reproductive organs were observed. 

Hexactiniae, in which the tentacles are arranged radially, more than 
one communicating with some or all of the endocoels. 


Stichodactylinae, with a marginal corona of tentacles, and iutermediate 
tentacles similar to those of the margin arranged iu radial series, each 
series consisting of from one to many tentacles. Musculature through- 
out weak ; no specially developed sphincter. 

This family was established by Hertwig ('82) for the reception of 
the two forms described by Moseley ('77) under the generic terra 


VOL. XVI,"| 
1893. J 



Corallimorpkus. In the " Supplement" Hertwig ('88) added to this 
genus, as another member of the family, the genus Coryndetis. Pre- 
vious to this, however, Andres ('83) had defined the family Cory- 
nactidfiB, splitting up Qosse's family GapneadsB, which he had previously- 
accepted ('80a), though recognizing that it was not altogether natural, 
and agreed with Hertwig in incorporating in his new family the genera 
Corynactis and dorallimorphus, adding also the genus Capneq. The 
name proposed by Andres is preferable to that of Hertwig, both on 
account of the greater autiquity of the genus, which serves as its sponsor, 
as well as on account of Hertwig's name carrying with it a significance 
which might give rise to misunderstaiidiug. Hertwig's name lias, how- 
ever, the priority in publicatiou, aiul it is therefore proper to retain it. 

Coralliinorphidae, with capitate tentacles, there being only one inter- 
mediate tentacle in each radial series; some of the marginal tentacles 
have no intermediate tentacles corresponding to them. 

No. 7315. Station 2839. Lat. 33-> 04' N.; long. 118^ 40' W. Depth, 414 fathoms. 
Two specimen's. ' 

These two specimens I found in a bottle which coutained also speci- 
mens of Myotianthus amhiguus and Pardctis vinosa. Both were in a 
very poor state of preservation, so that I can add nothiug to the ana- 
tomical description given by Hertwig ('82). One of the specimens 
was attached to a fragment of a gasteropod sbell. The columu measured 
1*™ in height and the disk had a diameter of 2.5議. There were uo 
indications of any tendency to iufold the margin, and sections demon- 
strated the absence of any sphiucter. 

The marginal tentacles were forty-eight in number, twelve being de- 
cidedly larger than the other thirty- six, and there were twelve inter- 
mediate tentacles corresponding to the larger marginal ones. The 
capitate natai e of the tentacles could be made out only with difficulty, 
but they certainly possessed that character. There appeared to be a 
slight thickening of the disc mesogloea at the base of each of the inter- 
mediate tentacles. 

StichodactyunjB with tentacles of only one form, short and tentacular, 
and covering the greater portion of the surface of the disc. Sphincter 
muscle strong and circumscribed, not embedded in the mesogloea. 


34. Corallimorphus profundus Moseley (1877) 




DiscosomidsB in which the colaina is not covered with tubercles, 
though verruca may be present in the upper part. Tentacles sbort and 

35. DiBCosoma foegiensis (Dana) M.-Edw. 
Plate xxxivj Figs. 113 and 114. 

SynonynM : Actinia fmegtensUy Dana (1846) ; Diicosoma fuegienns, Milne-EdwanU 
(1 お 7); Saga$tia fuegiem$isy Gosse (1860); Cereus faegiensis, Verrill (1868). 

Ko. 693. Station 2767. Lat. 4<P 03' 8. ; long. 58- 56' W. Depth, 52 fathoms. Four 
Bpecimens. | 

There is a certain amount of doubtfulness in this identification, since 
it is not possible to be certain as to whether the form described by Dana 
('46) is really a Discosoma. Milue-Bd wards ('57) considered it to De 
sucli, and Andres ('83) places it among the doabtfdl species of the 
same genus. So far as the description goes the AJhatroHs specimens, 
agree fairly well, and come from a station not especially remote from the 
locality in which Dana's form was found and from comparatively shal- 
low water. 

The four specimens differ considerably in size. The largest measures 
2.5cn» in height, and 4.5**"> iu diameter, whiJe the smallest is l*""" iu height, 
with a diameter of 2.5。《> at the base. Three of the i^pecimens are only 
partially contracted, the prominent lips of the mouth, and the outer 
cycles of tentacles being visible, while one of the smaller forms is com- 
pletely contracted, the tentacles and mouth being entirely concealed, 
and the body having the form of a cone, sloping gradually upward from 
the flat base. 

The base is adherent and has attached to it fragments of a brown 
cuticle. The mesogloea is thin and in some specimens has been rup 
tured, allowing the mesenterial filaments to protude. 

The ectoderm of the column has been macerated away for the most 
part, the few fragments that persist towards the linibus having a dingy 
white color in the preserved specimens, and presenting a reticulate 
appearance. The exposed mesogloea has a cream-white color, ana is 
smooth. In some of the specimens it has been considerably macerated, 
especially towards the upper part of tlie column, where the mesenteries 
are exposed. Owing to the absence of ectoderm, it is impossible to | 
determine whether or not verrucae may have been present in the upper 
part of the column. The sphincter muscle (PI. xxxiv, Fig. 113) is strong j 
and is of the circumscribed endodermal variety, resembling greatly 
that occurring in certain Bunodidae. 

The margin appears to have been lobed. The tentacles are numerous j 
and sbort, and are arranged in radial series. Their ectoderm is very 
ricniy supplied with neraatooysts. Their longitudinal mnsculature and 
the correspondiDg TOusculatiire of the disc is well developed, aud is not 
imbedded in the mesogkea. 

Digitized by 

1893. J 



The mouth is very prominent, and shows indistinct traces of a dark, 
slate-gray pigment. The mesoglcea of its lips is thickeued, the thick- 
ening gradually thiuniiig out, both towards the disc and towards tbe 
stomatodaBum. This is marked with longitudinal ridges, supported 
by mesogloeal elevations, and possesses deep siphonoglyphes. 

There are ninety-six pairs of mesenteries. Twelve of them, repre- 
senting the first two cycles, are perfect, the rest imperfect, the fifth 
cycle of forty-eight pairs being very small, hardly projecting above tbe 
endodenn. Reproductive organs are borne upon the forty-eight mes- 
enteries composing the third and fourth cycles. Ko acoutia are pres- 
ent. The loiigitudiual muscles of the mesenteries (PL xxxiv, Fig. 114) 
have a moderate degree of development, forming a rather diffuse pen- 
non. The parieto-baailar is, however, strong, forming a well-marked 
pouch upon the surface of the more developed mesenteries. 

Very decided differences exist between this form and D. anemone 
previously studied by me ('80), bat nevertheless a general similarity is 
well marked, showing itself in the sliape aud structure of the tenta- 
cles, the character of the sphincter muscle, and the deep siphonoglyplie. 
The musculature of the 】nesent せ ries lias, however, a very different 
ariaugement, and the relationship of the perfect end imperfect mesen- 
teries is quite different. These points, however may be justly regarded 
as specific. 

Tribe CERIANTHE^. Hert. 

Aothozoa, with a large number of unpaired mesenteries, and witli a 
single siphoncglyphe; the eight Edwardsian mesenteries are situated, 
four on each side, at the sulcar surface, and new mesenteries are added at 
the sulcular surface, being interposed, one on each side of the sagittal 
plane, between those immediately preceding them in time of formation. 
The base is not adherent and is usually. provided with a oore opening 
into tlie body-cavity. Column walls, with strong ectodermal mus- 

With the characters of the tribe. 

Genus CERIANTHUS, Delia Chiaje. 

Whether the form described below be correctly referable jbo the genus 
Cerianthus is questionable, inasmuch as it seems to differ in several 
particulars from any of the forms hitherto referred to the genus. An- 
dres ('83) divided the forms assignable to the family Cerianthidae into 
three genera (not including Arachnactis), but the characters upon which 
these genera were based hardly seem at present of sufficient importance 
to be considered generic. It seems to me preferable, at present, to 
assign the specimen described below to Delia Cliiaje's genera rather 
than to establish a new genus on insufficiently understood characters. 

Digitized by 


36. Cerianthus vaa, sp. uov. 

PL XXXIV, Figs. 117-119; PI. xxxv, Fig, 120. 

No. 726. Station 2838. Lat. 28。 12' N. ; long. 115^ 09' W. Depth, 44 fathoms. One 

The single Oeriantliid which I found in the collection gave so much 
promise of interesting results that I determined to sacrifice it to an 
anatomical investigation. Unfortunately, however, it did not prove to 
be well preserved, aud many points on which I had hoped to obtain 
definite information remained obscure, partly owing to the preservation 
and partly to the difficulties in tlie way of obtaiuiug all the necessary 
data from a single specimen. A Dortion of the upper part I cut in 
longitudinal sections in the endeavor to obtain, if possible, definite in- 
formation as to the absence of tentacles, aud was thus prevented from 
making a thorough study of the arrangement of the mesenteries. 

The specimen (PI. xxxiv, Fig. 117) measured 2.0"'1. in length and 
about 0.9。™. in diameter, and bad a decided vase-like appearauce. 
The margin was slightly reflected, and there was a distinct neck like 
constriction a little below it. The column was cylindrical, tapering 
gradually below, where there was a large, widely open pore communi- 
catiug with the interior cavity. The ectoderm had a pale brown color, 
and its musculature was richly developed in the manner characteristic 
of the Oeriaiithiclse. 

No tube accompanied the specimen, nor have I any information as to 
whether there was one when it was found. 

A remarkable peculiarity which attracted my attention at once was 
the apparent absence of tentacles. Neither at the margin nor upon the 
disc could any of these structures be found. It is possible tliat they 
may bave fallen away, an idea to which the fact that any sections 
through tbe margin did not show continuity of the column wall and 
disc, except in one or two cases, gives support. It seems hardly possi- 
ble, however, that if they bad been present they could have disappeared 
so completely as they seem to have done, and I am rather inclined to 
believe that they were absent or reduced to mere rudiments. 

The stomatodseum was narrow, extending only a short distance into 
the interior cavity. The portion wbich I used for Icmgitudinal sections 
probably contained the siphonoglyphc. Upon the other side of the 
stomatodaeum no siphonoglyphe occurred. 

In a section through the middle of the column (PI. xxxv, Fig. 120) 
twenty-two mesenteries could be counted. They showed a tendency to 
be arranged so that broad and narrow mesenteries "should alternate 
with one another, but this arrangement was frequently marred by a 
broad mesentery occurriDg in the place of a narrow one, and vice versa. 
It is evident, however, that two grades of mesenteries are represented 
in the section, one consisting of about twelve mesenteries quite wide 
and bearing reproductive organs as a rule, and one whose mesenteries 



I were much .narrower and were also destitute of reproductive organs. 

' Whether, as I am inclined to believe is the case, a third grade is pres- 
ent, extending only a short distance below the stomatodaeum, is uncer- 
tain. I was not able to prepare a satisfectory series of sections which 
would demonstrate this point, but sections made through a small por- 
tion of the column wall at a level with the stomatodseum seem to show 
a greater number of mesenteries tlian occur in a portion of the same 
size lower down. 

The character of the mesenteries attached to the sulcar directive I 
did not discover. Opposite the sulciilar end of the stomatodaeum I 
found a single mesentery (PI. xxxiv, Fig. 118, mea) which rapidly dimin, 
ished ill size as it passed backward, and even at the level of the lower 
edge of Ih^ stomatodaeum was reduced to the merest rudiment. This I 
take to be a newly formed mesentery, its fellow of the opposite side 
not having appeared. 

A decided abnormality was seen in sections taken about the middle 
of tbe column (PI. xxxv, Fig. 120), wliieli involved two mesenteries situ- 
ated at or near tbe sulcar surface (つ. These had united to form a baud 
from which two lamellse extended into tbe body cavity. A little higher 
these lamellae were likewise united so that a cavity was inclosed by the 
united mesenteries. 

I was not able to distinguish any acofitia in tbe region where they 
usually occur in Ceriaiitbids, though nieseiiterial tilameuts occurred on 
all the mesenteries. They were very imperfectly preserved, however, 
and did not allow of an accurate study. 

The reproductive organs arc bome by the widest mesenteries, which 
extend the greatest distance down the column. Both ova and sperma- 
tozoa seem to be borne by each Diesentery. (Plate xxxv, Fig. 119, ov and 
te.) Ova are certainly present and occurring with them, inclosed in the 
mesogloea, bodies which I take to be spermatozoa. They (te) vary much 
in size, occasionally being many times larger than the ova, and consist 
of a deeply staining wall crowded with small nuclei, a cavity occurring 
in the center of the larger ones aucl containing numerous nuclei, at- 
tached to which I could in some cases discover delicate appendages. 
They do not resemble the spermatozoa bundles of tbe Hexactiniae, but 
j bear a close resemblance to the testes of C, membranacem, figured by 
■ tbe Hertwigs. Cerianthns vas is accordingly most probably one of the 
I hermaphrodite Cerianthids. 

Tbe endoderm covering the mesenteries presents the same characters 
as that found in the same regions in C. amerimnuBy wbich I have de- 
scribed elsewhere ('90). 


No. 725. Station 2825. Lat. 24。 22' 15" N. ; long. IIO^ 19' 15〃 W. Depth, 7 fathoms. 
One specimen. 

No. 955. Station 2765. Lat. 36。 43' S; long. 56° 23' W. Depth, 11^ fathoms. One 


Part III. 


To anyone who has studied the habits of Actinians the dependence 
of the various species upon their surroundings is very evident. Some 
are to be found only on rocky shores, others prefer sandy bottoms, 
while others again make their homes on】y in 】nucl(ly flats. Some bury 
themselves in tlie sand or mud so that only tlie disk and tentacles pro- 
trude, others are to be found only on gasteropod shells inhabited by 
Hermit Crabs, while others again firmly clasp stems of Gorgonians. 
In other words, nearly every species has a more or less definite habitat. 

Fiirtlierinore, as a rule the various species have a more or less defi- 
nite distribution, so that it is possible to mark out more or less definite 
geographical regions characterized by tbeir Actinian fauna. Thus the 
eastern coast of the United States presents three fairly well defined 
regions so far as the Actinian fauna is concerned. North of Cape Cod 
we liave what may be termed the Boreal region, characterized by the 
occurrence, among other forms, of Tealia crassicornis^ Metridium mar- 
ginatum, and Cerianthm borealis Verr. Secondly, there is what Yerrill 
has called the Virginian region, which includes tlie Virginian and Caro- 
linian coasts, and probably Georgia to the south, and Delaware and 
part of New Jersey to the nortli, characterized by the presence ofPiy- 
mactis cavernata, Adamsia sol, and Cerianthus americanus among others; 
and lastly, there is the Florida region, characterized by forms identical 
with those of the West Indies. Northern New Jersey and Long Island 
Sound constitute an intermediate region possessing forms such as 
Metridium marginatum j reaching tbeir most perfect development m the 
Boreal region, and others, such as Moactis {Halcampa) producta and 
Paractis rapiformis, wbich belong properly to the Virginian region. 

When the distribution of genera is considered, however, this definite- 
ness, as might be expected, becomes more or less indistinct, though 
even with some of these distinct areas of delimitation can be estab- 
lished. With the larger groups tlie same liolds true, and even when 
the orders are considered a certain amount of limitation of their dis- 
tribution can be determined. The Actininae, it is true, bave a world- 
wide distribution, but, as I have pointed out ('89), tlie Sticliodacty- 
liiiae, though of wide distribution, have tlieir headquarters in the Pacific 
and West Indian regions, and it may be said in the regions of coral 

Our knowledge, however, of the Actinian fauna of a great deal of 
the Pacific and Indian Oceans and of the South Atlantic is as yet very 
slight, and it is hardly time to enter into an exhaustive discussion of 
the geographical distribution of the larger groups, families, and orders 
of the Actinaria. So far as the Albatross collection is concerned, there 

Digitized by 



is only one point that deserves special ineutioii in this connection, and 
that is. the very wide distribution which it reveals for certain deep-sea 

Actinauge verrilm and Actinauge faatigata have been obtained by 
the U. S. Fish Commission at various localities off* the eastern coast 
of the United States. The former is recorded from various stations 
from off the coast of i^^ova Scotia in the north to off Cape Hatteras in 
the south, from depths ranging from 30 to 506 fathoms. A. fastigata 
has been recorded from off Martha's Vineyard from a depth of 300 to 
980 fathoms. In the Albatross collection these forms were obtained 
from the following stations: 

A. verrilUi: Stations 2791, 2818, and 2839. 
A. fastigata: Station 2791. 

Station 2791 was off the coast of Chile; station 2818 off the coast of 
Ecaador, in the neighborhood of the Galapagos Islands; and station 
2839 off the coast of California, 

Another form, Actinostola callom, has likewise been obtained at 
various stations on the eastern coast, ranging from the Grand Banks 
of Newfoundland on the north ta Cape Fear, ! N". C, on the south, at 
depths varying from 50 to 640 fathoms. This form likewise occurs 
upon the west coast of America, having been obtained by the Alba- 
tross at stations 2792, 2807, and 2818, all of which are off the coast of 
Ecuador, and vary in depth from 392 to 812 fathoms. 

Since we liave seen that species of Actiniae are to a great extent de- 
pendent upon external conditions, this wide distribution of these 
deep-sea species is interesting. It seems improbable that they are 
wanting in the deep water of the southwestern Atlantic; or, in other 
words, that they occur sporadically upon the east and west coasts of 
America. Future observations will probably reveal their occurrence 
off the east coast of South America, a portion of the ocean whose 
Actiniau fauna is still to be studied, and it seems probable that they 
occur over the sea bottom of the western trough of the Atlantic 
throughout its entire extent, and doubling Cape Horn extend up the 
west coast in deep water as far north at least as California. Since we 
know that the temperature at considerable depths is fairly constant 
and low, it may be supposed that over this wide area these forms find 
conditions sufficiently similar, and have thus been enabled to extend 
their distribution. 


I give here in tabular form the localities and depths at which the 
various species of the Albatross collection were obtained : 

Note.— The species in italics are described for the first time in this report. 

Hertwig ('82), as the result of liis investigation of the Challenger 
Actiniaria, drew attention to two features presented by some deep-sea 
forms Avhich marked them out from those living in shallower water, 
jiamely the retrogression of the tentacles to stomidia and the uuusual 
arrangement of the mesenteries. The Albatross collection seems to 
lessen somewhat the importance of the first of these peculiarities by 
suggestiDg the propriety of doing away with the genera Polysiphonia 
and Liponema, but the second characteristic receives some support in 

Name of species. 

Station. Latitode. 



! 30 S. 
; 00 N. 

48 09 00 S. 
36 42 00 S. 


74 08 SOW. 
118 40 00 W. 

; 00 w. 
I 00 w. 

Abrolhos Islands. 
42 24 00 S. I 61 38 30 W. 
Pichilincne Bay. 
Straits ot Magellan. 
33 08 CON. 118 40 00 W. 

2783 I 51 02 30 S. 
t 06 00 S. 

34 00 S. 
t 08 00 
! 08 00 



74 08 SOW. 

70 40 36 W. 

63 00 00 W. 
118 40 00 W. 
118 40 00 W. 

8 16 30 N. 79 39 45 W. 
33 08 00 N. 118 40 00 W. 
Port Otwavi Patagonia. 
Lota, Chile. 
Galapagos Islands. 

38 08 00 S. 
37 00 S. 

51 34 00 S. 
45 22 00 8. 
33 08 00 N. 
33 08 00 N. 

48 09 00 S. 
36 42 00 S. 
36 47 00 S. 

8 44 00 N. 

1 03 00 N. 
38 08 00 S. 
33 08 00 N. 

L9 00 S. 
38 08 00 S. 
53 01 00 S. 

24 OU S. 

53 06 00 S. 

75 53 00 W 
81 00 00 W. 
87 06 00 W. 
89 54 30 W. 
65 46 00 W, 
68 00 00 W, 
64 20 00 W 
118 40 00 W, 
118 40 00 W. 

74 36 00 W, 
56 23 00 W. 
56 23 00 W, 

79 09 00 W. 

80 15 00 W. 

75 53 00 W, 
118 40 00 W, 

89 54 SOW, 
75 53 00 W, 
73 42 SOW, 
89 06 00 W. 

70 40 SOW. 

Pichiliiigu© Bay. 
36 47 00 S. 56 23 00 W. 

118 40 00 W. 
1 00 S. 58 56 00 W. 
1 00 N. 115 09 00 W. 


Tribe Edwardsiie : 

Edwardsia intermedia . 
Tribe Protactiniie : 

Oraetia JHomedw 

Tribe Hexactinue. 
Order Actininip. 

Family Halcampide : 

HcUcurias pUahu 

Peaehia Koreni 

Funily AntheadtP : 

Qia infecunda 

onia variabU 
Boloreridie : 

}cera occidua 

ra oecidtta 

% occidua 


ocera orevieomU 

Family Paractidw : 

Paractis lineolata 


Antholoba reticulata 

Antholoba reticulata 

Antholoba r«tieiilata 

Aetinemtu pUbrius 

Actinostola callosa 

ActiDostoIa callosa 

Actinostola calloga 

Actinostola exeeUa 

Actinostola exctUa 

Actinostola per^amentacea . 

Pycnanthut mali/onnut 


Family Sa^artid») : 

Sagartta lactea 

Sagartia Saneti Mathcei 

Sctgartia paradoxa 

Sacartia crispata 

Adamsia (f) involvent 

Actinange Verrillii. 
Aetinauge Verrillii. 
Actinauge Verrillii. - 
Actinange fast igata . 
Chitonanthiis j>et'tinatii8 
Stephanactu kyalonematis . 

Family Bunodidfe : 
ijeMteali^ badia 
Family Phyllactulse . 
Omactis cali/omica . 
CradaetU dxgitata... 
tichodactylii.fe : 
Family CoraliiniorphidFB : 


Order : 
ally Con , 
Corallimorphns profundus. 

Family DiscosomidsB : 

Di8CO80m a fuegieosiR - 
Tribe Ceriantheie : 

Cerianthus vas 

1 2 34 5«789 w wu^ wn s SMS お^? 5 s5 js お % 

^^^^^^^^^ 認 ^ 



Digitized by 


1893. J 



the discovery of OractiSj although tbe Faractinice have been removed 
from their high place. Two out of three of the genera forming the 
Tribe Protactiwiae are deep-sea forms, including under this head all 
those which live at depths approaching 500 fathoms. 

It is doubtful, however, if any such limitation can be set to distinguish 
deep-sea for shallow-water forms. What we mean by deep-sea forms 
are forms which live under conditions as a rule only to be found in the 
deeper water, one of the most important of which is perhaps great 
and constant cold. This is a condition which may be obtained at 
various depths according to latitude, and it is quite possible, in fact it 
does happen, that forms which in more southern latitudes are found at 
300 to 500 fathoms, may, in higher latitudes, occur at a depth of 30 to 
50 fathoms. If, however, a limit is to be given I should suggest one 
much less than that proposed by Prof, Her twig, perhaps as little as 100 
fathoms. It would be better probably to allow the limit to vary, con- 
sidering the zone at which the conditions are practically constant 
throughout the year to be the limit of true deep sea forms. 

There is definite evidence of a wide bathymetrical distribution of 
deep sea forms. For instance, Corallimorphus profundus was obtained 
by the Challenger from 1,375 to 2,025 fathoms, while the Albatross speci- 
mens were obtained from a depth of only 414 fathoms. So, too, we 
have seen that Actinostola callosa ranges from 50 to 812 fathoms, Acti- 
nmige fastigata from 300 to 980, and A. Verrillii from 30 to 677. Con- 
versely also shallow water forms may extend dowu to depths suflScieut 
to overlap the regions inhabited by what maybe considered deep-water 
forms. For instance, Antholoba reticulata is typically a littoral form, 
yet the Challenger obtained it from a depth of 55 fathoms, a depth 
greater than the highest limit from which either Actinostola calhsa or 
Actinauge Verrillii has been dredged. 

Making allowance for such eases, however, it is not difficult to divide 
the Actiniaria into such forms as are typically deep-sea dwellers and 
those which inhabit shallower waters. Reviewing the various families 
as to their peculiarities in this respect, it will be found that certain 
groups may be assigned to one or other category, while others have 
representatives in both. Among these latter are the Edwardsise, Pro- 
tactinii©, Sagartidae, Paractidae and Oorallimorphidae ; among the 
SagartidaB the Sagartinae are principally shallow-water forms, though 
Bome such as Sagartia lactea and Adamsia ( ?) involvens occur in deep water, 
while the ChondractininfB are essentially deep-water forms, though Phel- 
lia has several species dwelling in the littoral zone. The Paractiaae, too, 
though containing littoral forms are apparently more abundantly rep- 
resented in deep water, and it is interesting to notice that in these as 
well as in the C hondraeti iiintB, the deep-water forms are characterized 
by the thickness and firmness of the mesogloea of the column walls. 
The Boloceridfe so far as known are deep-wiiter forms, as are also the 
genera Polystomidium ^ PolyopiSj and Sicyonis; and, on the other hand, the 
Antheadae, Biinodidae, Phyllactidte, Heteractidae^ Thalassiaiithidse, and 


in fact all the forms with abnormally-shaped tentacles, excluding those in 
which these structures are reduced to stomidia, are essentially inliab- 
itaiits of shallow water. Perhaps an explanation ot the development 
of fronds as in the Phyllactidae and of branching and nodulose tentacles 
in shallow-water forms may be found iu the greater or less mimicry 
of tbe plant forms, with which littoral actinians are associated, which 
is thus produced, and which would serve as a protection from carniv- 
orous enemies. 

As with the geographical distribution, however, much yet remains to 
be done before any proper generalizations as to the significance of and 
the causes which govern the bathymetrical distribution of the Actini- 
aria can be made, and the remarks here presented are simply a sketchy 
outline of some of the ideas that have suggested themselves during the 
investigation of the Albatross collection. 

May, 1892. 

Since the preceding report was completed I have had the opportunity 
of examining the collections of Actinians iu the museum at Berlin, 
and also the Challenger collection in the Natural History Department 
of the British Museum, and must express my sincere thanks at this 
earliest opportunity to Prof, von Martens and Prof. Jeffrey Bell for the 
courtesy with which they acceded to my request to examine these very 
valuable collections and for their great kindness in affording me every 
facility for studying them. I also desire to state iny obligations to my 
friend Prof. A. C. Haddon for many valuable suggestions and much 
interesting information with regard to the European Chondractinitue, 
as well as for the opportunity of examining the valuable collection of 
forms belonging to that group which he possesses. 

As the result of my studies of these collections I have been able to 
confirm the correctness of certain suggestions made in the report, and 
also have obtained new light upon tlie identification of certain forms, 
and have thought it advisable to incorporate in this Report in the form 
of an appendix some of the more important of my results. 

Anemonia variabilis (p. 147). 一 In the Berlin Museum are preserved the 
forms described by Studer (,78), which were collected by the Gazelle 
expedition, and among them is a form which seems to be identical with 
that described above as Anemonia variahim. This is Corynactis camea, 
Studer. In size and habitat it agrees very closely with the Albatross 
specimens, and the capitate character which Studer describes for the 
tentacles is not at all well pronounced. It was upon this character 
that Studer relied in assigning it to the genus CorynactiSj but the ten- 
tacles are plainly arrangea m cycles, a- fact whieli inaybfj deduced from 
his statement that the tentacles are "z^hlreich iu zwei Eeihen." The 
similarity is so striking that, taking it into consideration wjtU the fact 
that both Ixave tlxe same habit, and come from esseutially the same 
locality aucl depth, I have no hesitation in pronouncing for its identity 


VOL. XVI, 1 
1893. J 



with the Albatross specimens, whose name should consequently be 
changed to Aiiemonia carnea (Studer). 

Boloeera hrevicornis (p. 158). 一 lu the Report I have expressed my 
opinion that Hertwig's Liponema miiltiporum is a Boloeera from which 
all the tentacles have fallen away. After an examination of the Chal- 
lenger specimens I feel more than ever convinced that such is the case. 
It is, however, as I suspected, specifically distinct from B. hremcornis. 

Gieniis Actinerntis (p. 165). 一 There can be no doubt but that Hertwig's 
Folysiphonia tuber osa properly belongs to the genus Actitiernus, though 

! its specific distinctness from all forms of that genus liitlierto described 
is exceedingly probable. 
It is interesting to note iii connection with the cxteusiou of the 
geographical range of the genus from the western basin of tbe Atlantic 
I to the Pacific that Iladdou * has recently noted its occurrence in 

i the eastern portion of the Atlantic, iu 750 fathoms oft' the southwest 
coast of Ireland. 

Actinostola callosa (p. 167). 一 Hertwig's Dysactis crmsicornis is un- 
doubtedly identical with this form. Tbe description given by Verrill 
I of Urticina callosa was published iii '82, and Hertwig's report of the 
Challenger Actiniaria appeared iu the same year, as did also a preliminary 
report, t It consequently is a question as to which name lias the 
priority. There can be no question as to validity of VerrilPs generic 
term, and it seems probable that his original description, which appeared 
in the March-April number of Sillinian's American Journal of Science, 
slightly antedates Hertwig's preliminary report. Leaving this aside, 
however, it seems preferable to adopt Verrill's name in its entirety, 
since the term crassicornis has a prior association with a member of 
the genus Tealia. 

j Genus Chitonanthus (p. 189). ― In establishing this genus I have laid 
stress upon two features : tbo presence of capitular ridges aud the ab- 
1 sence of bulbous enlargements at the bases of tbe tentacles. The im- 
I satisfactory nature of the classification of the ChondractininjB alluded 
I to above is priucipaUy due to the importance bestowed upon the nature 
and arrangement of the tubercles. The specimens of Chitonanthus pec- 
tinatus in tbe Albatross show of liow little iiiipoitauce this feature may 
! be in soinp cases, and it seems advisable to seek for some more constant 
characters. It is possible tbat these are to bo found iu the nature of 
the capitulum aiul of the bases of the tentarles. * The genus Actinaiige: 
seems well marked off, but this is not the case with Chondractinia^ Chit- 
onactiSj aud Mormathia, genera established principally on tbe nature of 
1 the tubercles, or on their ariangemeiir. It is hoc improbable tbat it 
will prove necessary to fuse these genera into one, removing from it,, 
towever, Hertwig's Chitonanthus {Phellia) pectinatus and Haddon's 

*A. C. Haddon.— Report on tlio Actinia dredged off the southwest coast of Ireland, 
m May, 1888. Proc. Roy. Irish Acad., 3(1 ser., Vol. i, 1890. 
tSitzungsber. Jenaiscb. Gesellscli, 1882. 

Proc. M. 93 —— 14 


Hormatnta Andersoniy both of which have been referred by the latter 
author to the genus Harmathia, 

The Ghondractininae^ if this suggestion prove worthy of acceptance, 
would then consist of the genns Hormathia characterized by the pres- 
ence of tubercles and a smooth capitalum, and by the absence of bulb- 
ous enlargements at the bases of the tentacles; the genus Actinauge 
possessing tubercles, a ridged capitulmn, and bulbous enlargements to 
the tentacles; Chitonanthus . with tubercles, and capitular ridges, hnt 
without tentacular bulbs; and Stephanactis, if it prove to be a "good" 
genus, without tubercles, capitular ridges, or tentacular bulbs, but with 
a clasping base. To these it may be necessary to add Phellia without 
tubercles, capitular ridges, or tentacular bulbs and without a clasping 

November 8, 1892. 


'80. Andres, A. Intorno alV EdwarcMa ClaparMii (Halcampa Claparedii Pane.). Atti 

R. Accad. Lincei in Roma. Vol. v. 1880. 
'80a. Andres, A. Prodramua neapolitance aciiniarum faunae, addito generalis etctiniarum 

bibliographi€e catalogo, Mitth. a. d. Zool. 8tat. zu Neapel. Bd. ii. 1880. 
'83. Andres, A. Le Attinie, Atti R. Acoad. Lincei. Vol. xiv. 1883. 
,91 . Van Beneden, E. Recherches «wr le d^veloppemen t dea ArachnacHs, Contribution d la 

morphologie des CManthides, Arch, de Biol. T. xi. 1891. 
,88. Blochmann, F., and Hilger, C. TJeber Gonactima proUfera Sars, eine durck Quer- 

theilung 8ich vermehrende Actinie, Morph. Jabrb. Bd. xm. 1888. 
'90. Boveri, Th. Ueber Entwicklung und Verwandtschafisbeziekungen der Aktinien. 

Zeit. fur wiss. Zool. Bd. xlix. 1890. 
'91. Carlgren, O. Beitrage zur Kenntniss der Aetinien-Gattung Bolocera Goaae, (Vor- 

liinfigo MitteiluDg). Ofvers. KoDgl. Vetens. - Akadem . Forhandl. Stock- 
holm. 1891. 

,46. Dana, J. D. Zoophytes of the U.' S. Exploring Expedition of 1838-^42. 1846. 

(The Atlas was published in 1849.) 
'46a. Dana, J. D. Structure and ClaasificaUon of Zoophytes. Philadelphia. 1846. 
W. Danielssen, D. C. Actinida of the Norwegian North Atlantic Expedition. 

Christiania. 1890. 

'88. Dixon, A. F. On the arrangement of the mesenteries in the genua Sagartia (Gosse). 

Sci. Proc. Roy. Dublin Soc. N. S. Vol. vi. 1888. 
,88. Dixon, G. Y, Bemarka on Sagartia venusta and Sagartia nivea. Scient. Proc. 

Roy. Dublin Soc. N. S. Vol. vi. 1888. 
Dixon, G. Y. & A. F. Notes on Bunodes thallia, BuDodes verrucosas, and 

Tealia crassicorais. Scient. Proc. Roy. Dublin Soc. N. S. Vol. vi. 18S9. 
,60, Duchassaing &, Michelotti. M4mo%re sur les Coralliaires des Antilles, Mem. 

Reale Accad. di Torino. 2'"« s^rie. T. xix. 1860. 
,66. Duchassaing & Michelotti. Supplement au Memoire sur les Coralliaires des 

Aniilles, Mem. Reale Accad. di Torino. 2'"« s^rie. T. xxiii. 1866. 
,34. Ehrenberg, C. G. Die Corallenthiere des roihm Meeres physiologisch «nier9$ickt und 

systematisoh verzeicknet. Berlin. 1834. 
,54. Gay, C. Hiatoi'ia ftsica y polttica de Chile, Zoologie. T. vni. Paris. 1854. 
,55. Gosse, P. H. On Peachia haatata with observationa on the family of Jctiniadai, 

Trans. LinnsBan Soc. Vol. xxi. 1855. ' 
'58. Gosse, P. H. On the British ActinuB. Ann. and Mag. Nat. Hist. Srdser. I. 1858. 
'60. Gosse, P. H. Actinologia Britannica, A history of the British 8ea-AmemoMe$ and 

Corah* LondoiL. 1860. 

Digitized by 

VOL. XVI, , 

1893. J 



'87. Hftddon, A. C. Note on ike arrangement of the mesenieries in the parasitic larra of 
Halcampa chryaantkellum (Peach). Scient. Proc. Roy. Dublin Soc. N. S. 
Vol. V. 1887. 

'88. H addon. A. C. On two species oj Aciiniw from the Mergui Archipelago, collected for 
the Trustees of the Indian Museum, CalcuUa, bif Dr, John Anderson, F. R, 5., 
Superintendent of the Museum, Jouru. LinnsBan Soc. Zool. Vol. xxi. 1888. 

'89. Haddon, A. C. A revision of the British Actiniae. Part I, ScientT. Trans. Royal 
Dublin Soc. Ser. II. Vol. iv. 1889. 

77. Von Heider, A. Sagartia troglodytes, Gosse, eiu Beitrag zur Anatomie der Actinien. 

Sitzungsber. d. k. Akad. d. Wissensch.Wien. Matb.-uat. Classe. Bd. Lxxv. 

'79. Hertwig, O. & R. Die ActiMie» anatomiach und hUiologisch mit besonderer 

BeruckaichiiguMg des NervenmuakeUifatema Hutersucht, Jena, 1879. 
'82. Hertwig, R. Report on the Actiniaria dredged by H, M. S. Challenger during the 

years 1873-1876, Rep. on the sclent, results of the voyage of H. M. S. 

Challenger during the years 1873-76. Zoology. Vol. Yi. 1882. 
'88. Hertwig, R. Supplement to Report on the Acimiaria dredged by H, M, S, Challenger 

during the years 1S73-1876, Report on the scient. results of the voyage of 

H. M. S. Cballenger. Zool. Vol. xxvi. 1888. 
'61. Johnson, J. Y. Notes on the Sea-Anemones of Madeira. Proc. Zool. Soc. London. 


'32. Johnston, G. Illustrationa in British ZMogy, Actinia Tuedio'., Loudon's Mag. 

Nat. Hist. Vol. v. 1832. (This paper I have not seen.) 
'47. Johnston, G, History of the British Zoophytes, London. 1847. (This work I 

have not seen.) 

'80. Jourdan, £. Kechervhes zoologiquea et histologiques sur Its Zoantkairea du Golfe de 
Marseille, Ami. des Sci. Nat. 6"'。 b6t, T. x. 1880. 

78. von Koch, G. Mitth^luttgeH Uber Cwlenteraten. Gcphyra Dohrnii. Morpli. Jahrb. 

IV. 1878. 

,56. Koren 6l Danielssen. Sye Jctinier Siphouactinia et AcUnopsis. Fauna litt. Nor- 
vegite. P. II. 1856. 

77. Kluuzinger, C.B. Die CoraUthiere den rothen Meeres. 1^^ The": Die Alcyonaricn 

und Malacodermen. Bed in. 1877. 
'30. Lesson, R. P. Zoologie, Voyage aiituur cUi moiide sur la Corvette de S. M. la 

Coquille pendant les aundes 1822-,25. Paris. 1838 et seq. 
,17. Lesneiir, C. A. Obaerrationa on several species of ^Ictinia, Jouru. Acad. Nat. Sci. 

Philadelpbia. Vol. i. 1817. 
'60. Liitken, C. Nogle Bemdrkinger om de ved de danske Kyster iagttagne Arter af AcU- 

niernes Gruppe. Naturhist. Foreu. Videusk. Meddelelsen. 1860. (This 

paper I know only through tbo reference to it by Haddon, 1889.) 
'89. McMurricli, J. P. The Actiniaria of the Bahama Islands, W, /. Jouru. of 

Morph. Vol. III. 1889. 
'89a. McMurrich, J. P. A contribution to the Actinology of ike Bermutias. Proc. 

Acad. Nat. Sci. Philadelphia. 1889. • 
,90. McMarrich; J. P. Contributions on the Morphology of the Actinozoa. I. Th. 
• structure of Cerianthns Jmericauns. Joiirn. of Morph. iv. 1890. 

'91. MeMurrtchy J. P. Contributions on the Morphology of the ActtHozoa. II, On the 

development of the Hexaciinice. Joarn. of Morph. iv. 1891. 
'9la. McMurrich, J. P. Contrihuiiona on the Morphology of ike Actinozoa. III. On 

the phylogeny of the Actinozoa, Journ. of Morph. v. 1891. 
'82. Marion, A. F. Aciinaires AtlantUiues des dragages de Vaviso " le Travailleur" 

Comptes RenduB. T. xciv. 1882. 
'57. Milne-fcdwarda, H. Histoire natitrMe de» ComlUmrea ou polypes proprement dits. 

T. 鳳. Paris. 1857- 

Digitized by 


,tJ. Mofteley, H. N. On new forms of Actiniaria dredged in the deep sea; with a de- 
scription of certain Pelagic aurf ace-swimming species. Trans. Linusean Soc. 
2me ser. Zool. Vol. i. 1877. 

'78. Mosely, H. N. On the structure of the Stylaslerida'. Phil. Trans. Roy. Soc. 1878. 

,41. de Quatrefages, A. M4nioire eur lea Edicardaiea {Edwardsia Nob.), nouveau genre 
de la famille des Actinies, Ann. Sci. Nat., 2me ま T. xvii; 1841. 

,29. Rapp, W. Ueher die Poly pen im Allgemeinen und die Actinien insbesondere, 
Weimar. 1829. 

,81. Ridley, S. O. Account of the Caslenterata collected during the survey of H. M, 8, 
Alert m the Straits of Magellan and on ihe coast of Patagonia. Proc. Zool. 
Soc. London. 1881. 

'78. Stiider, T. Anthozoa polyaciiniaf toelche wdhrena aer Reise S, M, Corvette Gazelle 
um die Erde gesammeli wurden. Monatsber. Akad. Wissensch. Berlin. 1878. 

,64. Verrill, A. E. Revision of the Polypi of the Eastern Coast of the United States. 
Mem. Boston Soc. Nat. Hist. Vol. i. 1866-'69. 

'65. Verrill, A. E. Classification of Polyps: (Extract condensed from a Synopsis of the 
Polypi of ihe North Pacific Exploring Erpedition, under Captains Binggold 
and Bodgersy U. S, N.) Proc. Essex lust. Vol. iv. 1865. 

,67. Verrill, A. E. Synopsis of the Polyps and Corah of the Xorih Pacific Exploring 
Expedition, under Commodore C. Ringgold and Capt. John. Rodgerpj U. S. N" 
from 1853 to 1856, Collected hy Dr, Wm, Stimpsony Naturalist to ihe Expe- 
dition. Proc. Essex Inst. Vol. v. 1867. 

68. Verrill, A. E. Review of the Corah and Polyps of the West Coast of America, 
Trails. Connecticut Acad. Arts and Sci. Vol. i. 186 く 

,73. Verrill, A. E. Exploration of Casco Bay hy ihe U, S. Fish Commission in 
Proc. Amer. Assoc. Adv. Sci. Vol. xxii. 1873. 

'79. Verrill, A. E. Notice of recent additions to the marine fauna of the Eastern Coast of 
N. America. Amor. Jonrn. Sci. and Art. ser. Vol. xvii. 1879. 

,82. Verrill, A. E. Brief contributions to Zoology from Ihe Museum of Yale Cottage. No. 

L. Notice of the remarkable fauna occupyinfj ihe outer banks off ihe aouthef-n 
coast of New England, Amer. Jouru. Sci. 3"* ser. Vol. xxiii. 1882. 

,83. Verrill, A. E. Report on the Anihozoa, and on some additional species dredged by 
ihe ''Blake*' in 1877-1879, and hy )hc U. S. Fish Commiasion Steamer " Fiah 
Hawk" inmO-,S2. Bull. Miis. Comp. Zool. at Harvard Coll. Vol. xi. 1883. 

,88. Wilson, H. V. On ihe development of Manicina areolata. Journ. of Morpb. Vol* 
II. 1888. ' 

,90. Wilson, II. V. On a new Acfinia, Hoplophoria coraUigena. Studies from the 
Biol. Lab. Johns Hopkins Univ. Vol iv. 1890. 




m ― 



basal muscle. 

mes ― 




m/ , 

muscle fibres. 


column wall. 

m び : 



coronal tubercle. 

neni = 



capitular ridges. 

ov ― 




pbm - 

parieto-basilar muscle. 


directive mesenteries. 

sph = 

sphincter muscle. 



8t = 




t 二 




te = 


foreign incmstatioiiH on the col- 

tm = 

transverse muscles. 

umn wall. 

isp = 

sphincter of tentacle. 

]ougitudinal muscles. 

lu = 


Digitized by 



Plate xix. 

Fig. 1. Edwardsia intermedia. Nat. size. 

タ.. Transverse section through introverted Bcapns of E, intermedia. Zeiss A 2. 
S. Transverse section through mesentery of E. intermedia, Zeiss A 2. 

4. Transverse section of column wall of E. intermedia, passing through a tuber- 

cle. Zeiss D 2. 

5. Oractis diomedew, viewed from the side. Nat. size. 

6. Oraeiia diomedece, viewed from above. Nat. size. 

7. Transverse section of tentacle of O. diomedece near its base. Zeiss A 2. 

8. Transverse section of column of 0. diomedew. The roman numerals indicate . 

the probable embryological successiou of the mesenteries. (An error has 
been made in tbe reproduction of tbis figure. The mesentery numbered 
V should be I, and the imperfect meseutery intervening between this and 
III should be V.) 

J^g, 9, Transverse section of a perfect mesentery of O. diomedece. Zeiss a 2. 

JO, Transverse section cuttiug columu wall, disc aiid base of the tentacles of O. 

diomedew. Zeiss A 2. 
11. Longitudiual section tbroagh margiu aud upper part of the column wall of O. 

diomedew. Zeiss a 2. 
Jg. Halcurias pilalus. Nat. size. 

13. Transverse section through mesentery of H. pilatus, Zeiss A 2. 

FHa, 14, Transverse section through the upper part of tbe columu of H. pilatus. a 2. 

15. Transverse section through ectoderm of the column wall of H. pilatut. 

Zeiss D 2. 

16, Peachia Korenu Nat. size. 

17. Transverse section through a mesenterial filament of Actinia infecunda. 

Zeiss D 2. 

18. Anemonia rariahilia, Nat. size. 

19, Transverse section through the column of A. variabilis, x and y deuute areas 

where the regular sequence of perfect aud imperfect mesenteries is iuter- 
fered with, x about 10. 

20. Condylactis cruentata. Nat. size. 

21, Transverse section of primary mesentery of Condylactis cruentata, Zeiss a 2. 

22, Myonanihus ambigum, Nat. size. 

Pig, £S, Transverse section of sphincter muscle o{ M, ambiguus, 

$4, Transverse section of sphincter muscle of Bolocera occidua. Zeiss a 2. 

► タ 5. Portion of transverse section of primary mesentery of B, occidua. Zeiss A 2. 
S6, Portion of transverse section of a tentacle of B. occidua. Zeiss a 2. 

27. Basal portion of longitudinal section through a tentacle of B, occidua. Zeiss 

A 2 unscrewed. 

28. Portion of transverse section of a tentacle of B.pannoaa. Zeiss A 2. 

29. Transverse section through a perfect mesentery of B, pannosa. Zeiss A 2. 

Fig. SO. Transverse section of the sphincter muscle of B, pannosa. Zeiss A 2. 
SI. Transverse section of the sphincter muscle of B. hrevicornis. Zeiss a 2. 

Outer portion of transverse section of meseutery of B, hrevicornis. Zeiss A 2. 

Plate xx. 

Plate xxi. 

Plate xxii. 



^W' S3. Inner portion of transverse section of mesentery of brevicornia. Zeiss A2. 
S4, ParacUs lineolata, Nat. size. 

35, Transverse section of imperfect mesentery of Paraclis lineolata,^ Zeiss a 2. 

36, Transverse section of sphincter muscle of P. lineolata. Zeiss a 2. 

37, ParacUs vinosa. Nat. size. 

S8. Transverse section of tentacle of F. vinosa. Zeiss A. 2. 

39. Portion of transverse section of mesentery of P. vinosa, Zeiss A 2. 

40. Portion of transverse section of sphincter muscle of P. vinoaa, Zeiss D 2 

Plate xxiv. 

Fig. 41. Transverse section of sphiucter of P. vinosa. Zeiss a 2. , • 

42. Transverse section (somewhat oblique) of base of a tentacle of Aciinernua 

plebeiufi. Zeiss A 2. 

43. Portion of transverse section of sphincter muscle of A. pleheius, Zeiss A 2. 

44. Portion of transverse section of primary mesentery of A. pleheius. Zeiss A 2. 

45. Transverse section of mesentery of A. plebeius. Zeiss a 2. 

46. Section, partly diagraiomatic, showing the arrangement of the mesenteries 

of Actinostola callosa. 

Plate xxv. 

Fig. 47. Actinostola callosa. Nat. size. 

48. Transverse section of a tentacle of A. callosa. Zeiss A 2. 

49 and 50. Transverse section of a perfect mesentery of A. callosa. Zeiss A 2. 

51. Transverse section of sphincter muscle of callosa. x 2. 

62. Portion of transverse section of sphiucter muscle of A. callosa. Zeiss A 2. 

Plate xxvi. 

Fig. S3. Actinosiola exceUa. Nat. size. 

54. Transverse section of sphincter muscle of A. exceUa. Zeiss a 2. 

55. Transverse section of perfect mesentery of A. exceha. Zeiss a 2. 

56. Transverse section of portion of a tentacle of ん exceUa. Zeiss a 2. 
67. Actinosiola pergamentacea. Nat. size. , 

58. Tangential section of portion of the disk of a specimen of Actinostola per go- 
mentacea. Zeiss A 2. 

Plate xxvii. 

Fiff.$9. Transverse section of sphincter muscle of A. pergamenfacea. Zeiss a 2. 
60. Portion of transverse section of a tentacle of A. pergamentacea. Zeiss A 2. 
- 61. Portion of transverse section of mesentery of 丄 pergamentacea. Zeiss D 2. 

62. Outer portion of transverse section of mesentery of A. pergammtacea above 

the level of the parieto-basilar muscle. Zeiss a 2. 

63. Middle portion of same section as that from which Fig. 4 was drawn. 

64. Pycnanthus maliformia. Nat. size. 

65. Transverse sectiou of sphiucter muscle of P, maliformis. x 4. 

66. Portion of transverse sectiou of sphincter muscle of P. maliformis from the 

region indicated in Fig. 65. Zeiss D 2. 

67. Transverse section through a perfect mesentery of P. maliformis. Zeiss a 2. 

Plate xxviii. 

Fig, 63. Tangential section through disk of a specimen of P. maliformis. Zeiss A 2. 

69. Cymhactis fceculenia, Nat. size. 

70. Transverse sectiou of sphincter muscle of C. faeculenta. Zeiss a 2. 

71. Transverse section of mesentery of first cycle of C. fceculenta. Zeiss a 2. 

72. Sagartia lactea. Nat. size. 

Digitized by 



Fig. 73. Transverse section of sphincter muscle of S. lactea, x abont 2. 

74. Transverse section of of primary mesentery of S, lactea, Zeiss a 2. 
76. Transverse section of acontium of S* lactea. Zeiss D 2. 

76. Portion of transverse section of sphincter muscle of S. laciea, Zeiss A 2. 

77. Sagartia SancH Mathwi. Nat. size. 

78. Transvenie section of sphincter of S, SancH MathctL 

Zeiss A 2. 
, 79, Sagartia paradoxa, Nat. size. 
* 80. Transverse section of spliiucter muscle of S, paradoxa, Zeiss A 2. 
81. Transverse sectiou of a directive mesentery of S. paradoxa, 
82 and 83. AdaiMia f inrolvena, Nat. size. 

Fig. 84. Semi diagrammatic section through the column of S, paradoxa, showing 
the arraugement of the mesenteries. The roman uumerals indicate the 
cycles of mesenteries, x and y the abnormal pairs. 

85. Transverse section of sphincter muscle of A, involvena, 

86. Trausverse sectiou of sphincter muscle of Actinauge VernlUi, Zeiss A 2. 

87. Transverse section through tbn upper third of the sphincter muscle of A, 

VerrilUL Zeiss A 2. 

88. Transverse section through the lower third of the sphincter muscle of A. 

rerriUii. Zeiss A 2. 

89. Portion of upper part of column of J. VerrilUi, tlie specimeu having been 

divided lougitiidiually. Nat. size. 

Fig, 90. Transverse section of mesentery of the second cycle of A. Fernllii, Zeiss A 2. 
91. Transverse sectiou through the base of a tentacle of A. Verrillii. Zeiii A 2. 
. 9^. Outer portion of transverse section of a mesentery of the first cycle of A. 
Verrillii. Zeiss A 2. 

93, Actinauge fastigaia. Nat. size. 

94, Transverse section of the spliiucter muscle of A, faaligata. 

95, Trausverse section of the lower part of the sphincter muscle of A, fasHgata, 

Zeiss A 2. 

96, Transverse sectiou of the upper part of the sphiucter muscle of A. faaiigata, 

Zeiss A 2. 

97, Transverse section of a mesentery of the first cycle of A . fasHgata, Zeiss A 2. 

F^ig. 98. ChitonanthuB peclinatus. Nat. size. 

99. View of surface of a dissected specimen of C. pecHnatua which had been 

divided longitudinally. Nat. size. 

100, Transverse sectiou of the sphincter muscle of C. pectinatua. x 4. 

101, Transverse section of portion of the sphincter muscle of C pectinaiue* 

Zeiss a 2. 

102. Transverse section of a mesentery of the first cycle of C. peclinatus. Zeiss 

a 2. 

10S» Stephanacti8 hyalonematis. Nat. size. 

104, Transverse section of sphincter muscle of Leiotealia hadia. Zeiss A 2 

105. Oulactis californica. Nat. size. 


Plate xxx. 


Plate xxxii. 


Plate xxxiii. 

Fig. 106. Transverse section of primary mesentery of L, badia. Zeiss A 2« 

107. Trans vBrse section of a mesentery of the first cycle of O. californica, 
lOS. Transverse section of sphincter muscle of O. californica. Zeiss A 2. 

109, Portion of margin of Cradactia digiinta, showing the tentacles and the 

fronds, x 2. 

110, View of frond of C. digiiata. Enlarged. 

111, Transverse section of sphincter muscle of C. digitata. Zeiss A 2. 

112, Transverse section of directive mesentery of C, digitata. Zeiss a 2. 

Fig. lis. Transverse section of sphincter muscle of Discosoma fuegiensis. Zeiss a 2, 

114. Transverse section of the raeseutery of tlie second cycle of Discosoma 

. f uegiensis. Zeiss a 2. 

115. Transverse section of sphincter muscle oi Anemoniaf inequalis. Zeiss a 2, 

116. Transverse section of a portion of the column of ^. inequalis. Zeiss a 2. 

117. Cei'ianthua vas. Nat. size. 

118* Dorsal portion of transverse section of the upper part of the column of C. 
vas, Zeiss a 2. 

119. Transverse section through gonoplioric region of a mesentery of C. vas, 
Zeiss D 2. 

Fig, IfSO, Transverse section through the middle of the column of C. tas. * points to 
the fusion of two mesenteries. 
121. Diagram showing the relation of the tentacles to the capitular ridges in 
Actinauge Verrillii. 

Plate xxxiv. 

Plat お xxxv. 




Digitized by 

Digitized by 





9-11. Oractis Diomedece. 12-13. Halcurias pilatus. 

Digitized by 

Digitized by 




s pi! ひ tu8. 16. Peach ia Koreni. 17. Actinia infecunda. 

18-19. Anemonia variabilis. 20-21. Condyhtciis cruentata. 22. Mi/onanthus ambUju us. 

Digitized by 


Digitized by 




Digitized by 

Digitized by 




36 '"""きんぼ ま 

39 40 

30. Bolocera pannosa. 31-33. B. brevicomis. 34-36. Paractis lineolata. 37-40. P. vinosa. 

Digitized by 


Digitized by 







Paractis vinosa. 42-45. Actinemus plebeius. 46. Actinostola callosa. 

Digitized by 


Digitized by 




47-52. Actinostola callosa. 

Digitized by 


Digitized by 




5 & -56. Actinostola excelsa. 57-58. A. peryanientacea. 


Digitized by 

Digitized by 





59-63. Actinostola pergamentacea. 64-67. Pycnanthus maliformis. 

Digitized by 


Digitized by 





68. Pycnanthus maliformis. 69-71. Cymbactis fceculenta. 72-75. Sagartia lactea. 

Digitized by 


Digitized by 



S. Sancti Mathoei. 
Adamsia (f) involvens. 

Digitized by 


Digitized by 






84. Sagartia paradoxa. 85. Adamsia (?) involvens. 86-89. Actinauge Verrillii. 

Digitized by 


Digitized by 




Digitized by 

Digitized by 




98-102. Chitonanthus pectinatus. 103. Stephanactis hyalonematis. 
104. Leiotealia baaia. 105. Outactis califoimica. 

Digitized by 


Digitized by 




106. Leiotealia badia. 107-108. Oulactia califomica. 109-112. Cradactis digitata. 

Digitized by 

Digitized by 




Digitized by 

Digitized by 




n n 


120. Cerianthus vas. 121. Actinauge Verrillii. 

Digitized by 

Digitized by 



Digitized by 


Digitized by 


3 2044 107 212 672 

Digitized by